初创公司尽调
尽调报告 robotics/hardware Series F 2026-05-19

Applied Intuition

自动驾驶与防务车辆的 Physical AI 基础设施

Applied Intuition 是自动驾驶开发基础设施里占主导地位的私营平台,深度打入 OEM,并在国防业务上扩张;基于现有证据,$15B 估值偏高,但若未披露收入能证实其声称的盈利能力和三位数增长轨迹,这个估值仍可辩护。

封面要素

最新估值 01
15 B USD [CO017]
累计融资(估算) 02
~$1.5B+ USD [CO018]
员工规模 03
1,000+ engineers [CO005]
OEM 客户 04
17 of top 20 global OEMs [CO007]
成立时间 05
2017 [CO001]

公司概况

Applied Intuition, Inc.(2017 年成立于加州 Sunnyvale)是一家私营技术公司,搭建自动驾驶和智能车辆所需的软件基础设施。其产品包括仿真工具(Simian)、数据管理平台(Spectral)、全栈 Self-Driving System(SDS)和 Vehicle OS,客户覆盖汽车 OEM、商用卡车公司和美国国防部。公司称已服务全球前 20 大汽车 OEM 中的 17 家,并确认美国陆军和空军为防务客户。Series F 轮由 BlackRock 和 Kleiner Perkins 共同领投,估值 $15 billion;此后 Applied Intuition 还宣布与 OpenAI 达成战略合作,把自己定位为 Physical AI 基础设施公司。

官网
appliedintuition.com
成立时间
2017-01-01
创始人
Qasar Younis, Balasubramanian Narayanan
创立地点
Sunnyvale, CA
总部
Sunnyvale, CA
产品
Applied Intuition 销售一套自动驾驶车辆开发工具,覆盖仿真、数据管理和安全验证,并捆绑用于量产部署的全栈 Self-Driving System(SDS)与 Vehicle OS。其 Physical AI 平台包含支持 MCP 的智能体接口,用于 AI 驱动的开发工作流。
客户
汽车 OEM(乘用车和卡车)、商用卡车公司、美国防务机构(陆军、空军)以及 AV / 出行初创公司。
商业模式
企业软件授权(仿真、数据管理、OS、SDS)、面向 OEM 集成的专业服务,以及防务项目合同。管理层称业务已盈利,且同比增速达到三位数百分比。
阶段
Series F (Private)
融资情况
2025 年 Series F 轮估值 $15B,由 BlackRock 和 Kleiner Perkins 共同领投,Fidelity 和 Lux Capital 参投。累计融资估计约 $1.5B+。
[CO001, CO002, CO003, CO004, CO005, CO006, CO007, CO017]

执行摘要

主要优势

  • 全球前 20 大汽车 OEM 中有 17 家使用 Applied Intuition 工具,形成深嵌入式分发护城河。
  • 已确认的 US Army 和 Air Force 国防客户带来粘性政府收入,也形成涉密技术护城河。
  • BlackRock 和 OpenAI 共同投资,表明市场押注的不只是 AV 仿真工具,而是实体 AI 基础设施。
  • 公司声称已盈利且同比三位数增长;若属实,Applied Intuition 就不同于那些烧钱却没有收入的 AV 同行。
  • 全栈产品(仿真 + 数据 + SDS + Vehicle OS)让公司比单点工具竞争者更深地接入 OEM。

主要风险

  • 收入和 ARR 未公开披露;没有财务数据,外部无法独立评估 $15B 估值。
  • $15B 估值需要 $500M+ ARR,或一个外部无法验证的国防 / 基础设施溢价倍数。
  • AV 赛道即便有 OEM 背书,也出现过重大失败(Argo AI 关闭、Embark 破产);行业时点风险仍在。
  • ITAR 和出口管制限制中国及部分国际市场 TAM;涉密软件也带来合规负担。
  • 开源 AV 仿真(CARLA)长期压制仿真工具授权收入的定价。

未决问题

  • 年经常性收入(ARR)和毛利率:估值承销里最重要的未知项。
  • OEM 合同深度:公司声称的 20 家中 17 家 OEM 关系,到底是生产部署还是试点 / PoC。
  • Series F 的确切融资金额、投后股权结构表和投资人治理权利。
  • DoD 合同载体和涉密项目价值:国防总收入无法从公开信息估算。
  • 竞争替代风险:公开证据没有显示任何 OEM 已从 Applied Intuition 转向竞争对手。

目录

Chapter 01

01公司概览

1.1 身份与使命

Applied Intuition, Inc. 由 Qasar Younis 和 Balasubramanian Narayanan 于 2017 年在加州 Sunnyvale 创立,目标是搭建开发自动驾驶和越来越软件定义化车辆所需的软件基础设施。公司现在把使命说得更宽:“推动世界运转的 Physical AI”。这表明它不想只被理解为仿真供应商,而是要成为汽车、卡车、防务系统乃至未来其他机器人平台的核心操作层。其产品栈围绕 Self-Driving System(SDS)、Vehicle OS 和车辆智能工具展开,反复强调仿真、测试、数据工作流和安全验证。这个定位重要,因为 Applied Intuition 站在价值链中杠杆很高的位置:它不是押注某一个车辆项目,而是销售可在多个项目、客户和车型之间复用的基础设施。 [CO001, CO002, CO003, CO004, CO040]

Applied Intuition 核心 KPI 快照
指标数值日期置信度来源 / 备注
估值$15B2025最新公布的 Series F 估值;具体轮次现金金额不如价格信号公开
累计融资~$1.5B+2026-05-19根据已披露轮次估计;精确累计金额仍是私有信息
工程团队规模1,000+ 名工程师2026-05-19当前公司官网和招聘材料
OEM 渗透全球前 20 大汽车 OEM 中的 17 家2026-05-19公司宣称的客户指标
办公地点覆盖北美、欧洲和亚洲 12+ 座城市2026-05-19About 和招聘页面列出全球办公室
收入 / ARR未披露2026-05-19unknown私营公司缺口;没有公开 ARR、收入或利润率披露
Embark 资产收购$71M 股权价值2023-08-02公司公告和 SEC 材料披露的上市公司交易

数据截至 2026 年 5 月,汇总自 Applied Intuition 公司页面、融资公告和部分第三方报道。收入、ARR、盈利能力和客户级收入结构未公开披露。

[CO005, CO006, CO017, CO018, CO031, CO043]
FO002: Applied Intuition 公司快照逻辑

身份、产品、客户采用和资本如何共同强化 Applied Intuition 的物理 AI 平台逻辑。

[CO003, CO004, CO029, CO038, CO040]

1.2 领导与治理

公司领导层仍带有很强的创始人印记。Qasar Younis 是 CEO,也是业务的公众面孔,履历横跨 Google 和 Y Combinator;Balasubramanian Narayanan 则作为联合创始人和创始 CTO,奠定了公司早期技术架构。公开材料显示 Peter Ludwig 现任 CTO,说明公司扩张后,工程组织已经从原始创始结构向外拓宽。Applied Intuition 还在 2020 年 Series D 轮把高知名度汽车业运营者拉入生态,包括前 GM CEO Rick Wagoner 和前 Daimler CEO Dieter Zetsche 担任顾问。这种组合给公司带来少见的汽车业信用,但没有消除治理不透明:公司仍是私营企业,不公布完整的当前董事会名单或股权结构图;在投资者关系、OEM 信任和战略方向上,它看起来仍明显依赖少数高级管理者。 [CO009, CO010, CO011, CO012, CO013, CO032]

领导层与创始人表
人物职务背景创始人 / 关键人物依赖度
Qasar YounisCEO、联合创始人前 Google 运营负责人和 YC 合伙人;公司外部形象代表创始人关键——投资人、OEM、招聘和战略关系都与他高度绑定
Balasubramanian Narayanan联合创始人、创始 CTO公司自动驾驶软件栈的创始技术架构师创始人高——原始产品和系统 DNA 仍与他相关
Peter LudwigCTO现任高级技术负责人,统筹规模化工程 / 产品组织关键高管高——随着产品范围超出创始团队,他对执行至关重要
Rick Wagoner顾问委员会Series D 前后加入的前 GM CEO关键顾问中——增强公司在 Detroit 和现有车企董事会层面的可信度
Dieter Zetsche顾问委员会Series D 前后加入的前 Daimler CEO关键顾问中——增加欧洲 OEM 可信度和战略建议

基于公司 About 页面、Series D 公告和独立资料来源整理。表格覆盖创始人、当前具名技术领导层和已公开披露的汽车行业顾问;私营公司完整董事会名单未公开。

[CO009, CO010, CO011, CO012, CO013, CO032]

1.3 融资历史与资本结构

Applied Intuition 的资本历程,从早期种子规模融资快速爬升到私营自动驾驶软件公司中估值最高的一批。公开可得证据指向:2017 年约 $2 million 的种子 / Series A,2018 年约 $40 million 的 Series B,2019 年 $125 million 的 Series C,以及 2020 年估值 $3.6 billion 的 $175 million Series D。公司随后披露了估值 $6 billion 的 $250 million Series E,之后又完成 Series F,估值 $15 billion,BlackRock 和 Kleiner Perkins 是领投方,Fidelity 和 Lux Capital 参投。投资者演进具有战略意义:Andreessen Horowitz、General Catalyst、Lux Capital、Elad Gil、BOND、Coatue 等早期风险投资者,如今旁边站着全球资产管理机构和蓝筹品牌投资者。累计融资额最好视为约 $1.5 billion 或以上的估计值,因为早期轮次的精确规模和最新一轮现金金额,在可取得公开来源中并未完整披露。 [CO014, CO015, CO016, CO017, CO018, CO030]

利益相关方或投资人图谱
投资人角色轮次战略重要性尽调问题
BlackRock领投 / 主要新增机构投资人Series F释放主流机构信心和潜在跨市场投资人兴趣确认持股比例、董事会 / 观察员权利以及任何老股出售预期
Kleiner Perkins领投 VC 投资人Series F增加顶级品牌背书和深厚企业网络支持确认持股规模、治理权利和后续跟投能力
Lux Capital多轮支持者Series C、Series E、后续轮参与者在公司从工具供应商上升为平台叙事的过程中持续支持按轮次梳理持股和任何按比例跟投保护
Andreessen Horowitz早期机构投资人Series A / Series C / 后续轮参与者早期品牌信号和硅谷网络杠杆确认持仓仍为新股为主,还是已有部分老股交易
General Catalyst早期机构投资人Series B / Series C / 后续轮参与者帮助建立早期企业级规模和可信度确认当前持股和任何董事会权利
Elad Gil成长期投资人Series D / Series E带来运营型投资人可信度和后期支持澄清持仓规模及其对未来融资的影响力
BOND / Mary Meeker成长期投资人Series E增加后期成长信号和市场叙事支持澄清持股、信息权和预期退出期限
Coatue成长期投资人Series D在估值跃升阶段增加公私市场跨界视角确认是否仍持有仓位,以及持有条款

投资人图谱反映各已公布轮次中公开具名的支持者。由于 Applied Intuition 是私营公司,未披露完整股权结构表比例、董事会席位分配或详细轮次条款,本图谱有意保持不完整。

[CO014, CO015, CO016, CO017, CO018, CO030]
FO003: Applied Intuition 快照 KPI

规模、融资、客户触达和尽调不透明度的简明快照。

[CO005, CO017, CO018, CO031, CO039, CO043]

1.4 业务里程碑

Applied Intuition 的里程碑显示,公司已经从 AV 开发工具扩展成更宽的自动驾驶基础设施平台。融资节点只是故事的一部分。PACCAR 选择其用于自动驾驶卡车开发,给公司带来一次重要的商用车验证;之后 Volvo Group 又提供了独立证据,公司也公开强调与 Volkswagen、Toyota、GM、Hyundai 的合作。2023 年 8 月,Applied Intuition 以 $71 million 股权价值收购 Embark 资产,在一个行业失败案例之后拿到困境但具有战略价值的卡车资产和数据。公司也一直强调防务动能,包括美国陆军项目,以及面向美国军事自动驾驶用例的更宽定位。OpenAI 合作是最近最清晰的战略信号:叠加最新融资,它表明 Applied Intuition 希望投资者和客户少把它看成点状工具供应商,多把它看成 Physical AI 部署的基础软件层。 [CO019, CO020, CO021, CO022, CO023, CO024]

里程碑表
日期事件类型金额 / 估值参与方含义
2017Applied Intuition 在 Sunnyvale 创立创立Qasar Younis;Balasubramanian Narayanan为自动驾驶和软件定义汽车建立软件基础设施路径
2017种子 / Series A 融资融资~$2MAndreessen Horowitz提供初始现金跑道,尽管详细经济条款披露很少
2018Series B 融资融资~$40MGeneral Catalyst支撑公司越过最早期阶段并扩大规模
2019Series C 融资融资$125M;累计融资约 $175M投资方包括 Lux Capital、Andreessen Horowitz、General Catalyst标志其成为 AV 开发工具类别领导者
2020Series D 融资与顾问委员会扩张融资$175M,估值 $3.6B投资方包括 Elad Gil、Addition、Coatue;顾问包括 Rick Wagoner、Dieter Zetsche估值显著跃升,同时提升汽车行业可信度
2020-06PACCAR 选择 Applied Intuition合作已公布商业合作PACCAR;Applied Intuition验证其适用于商用卡车场景
2022Volvo Group 宣布合作合作战略开发关系Volvo Group;Applied Intuition在全球卡车和工业车辆市场增加独立证据
2023-08-02收购 Embark 资产反向$71M 股权价值Applied Intuition;Embark Technology增加卡车资产和里程,同时凸显行业压力和整合
2024Series E 融资融资$250M,估值 $6B投资方包括 Lux Capital、Porsche Investments、Elad Gil、BOND、a16z、General Catalyst扩大资本基础,支撑公司越过核心仿真工具继续扩张
2025Series F 融资与 OpenAI 合作融资$15B 估值投资方包括 BlackRock、Kleiner Perkins、Fidelity、Lux Capital、OpenAI将公司重新框定为更宽的 physical-AI 基础设施
2025-2026国防进展因 U.S. Army 和更广泛军事定位而被强调合作客户 / 用例披露客户 / 相关方包括 U.S. Army、U.S. Air Force、Applied Intuition显示公司从汽车扩张到政府自主系统项目

该时间线聚焦截至 2026 年 5 月最关乎公司身份、融资、合作和反向背景的里程碑记录。早期轮次经济条款在私营公司披露不完整时按近似值处理。

[CO001, CO014, CO015, CO016, CO017, CO019]
FO001: Applied Intuition 公司里程碑时间线

从创立到当前物理 AI 定位的关键融资、合作、国防和收购里程碑。

[CO001, CO012, CO013, CO014, CO015, CO016]

1.5 规模与全球布局

当前公司快照的显著之处在于覆盖面很宽,尽管核心财务指标仍未公开。Applied Intuition 称拥有 1,000+ 名工程师,其中包括 40 名前 CTO 和 30 名前创始人,并在北美、欧洲和亚洲列出多个办公室,以支持跨国 OEM 项目。其客户页面称已进入全球前 20 大汽车 OEM 中的 17 家,公开引用还通过 PACCAR、Volvo Group 和美国陆军延伸到卡车和防务。合在一起,这些信号意味着公司在车辆软件栈中具备基础设施级的重要性。同时,公开证据仍留下重大尽调空白。收入、ARR、毛利率、客户集中度和盈利情况均未披露,这意味着估值承销过度依赖招聘密度、合作伙伴标识、投资者质量和里程碑公告等规模代理指标,而非经审计的经营表现。 [CO005, CO006, CO007, CO008, CO031, CO033]

1.6 展品

Chapter 02

02市场分析

2.1 市场规模与增长

对 Applied Intuition 核心市场来说,最干净的外部代理不是泛化的“自动驾驶”类别,而是自动驾驶车辆仿真与验证软件。MarketsandMarkets、Grand View Research、Allied Market Research、Mordor Intelligence 和 Business Research Insights 的口径汇总后,仿真细分市场大致聚集在 2024 年约 $2.5 billion 的规模,增长率从高个位数到高十几位数不等,2030-2035 年终点约 $7-8 billion。这个区间有方向性价值,因为 Applied Intuition 的产品栈围绕仿真、测试、验证和车辆软件工作流,而不是运营 robotaxi 车队或制造传感器。更宽的相邻市场大得多:Mordor 将 2024 年 ADAS 市场放在约 $33 billion,2030 年展望超过 $80 billion;MarketsandMarkets 预计自动驾驶软件到 2035 年约 $7 billion。 承销时,正确的规模测算应分层看。宽口径 TAM 可以合理合并 AV 仿真、自动驾驶软件、ADAS 相邻工具和防务自动驾驶预算,形成 $40-45 billion 的视角,但这会夸大近期可被外部工具消化的需求。更现实的 SAM 应围绕 OEM 项目、商用卡车和防务工具,接近 $10-12 billion。汽车 OEM 工具大概率占可服务池的大头,因为即便 Level 4 全面部署继续推迟,车企也必须验证软件定义车辆和越来越高级的驾驶辅助。第三方自动驾驶工具供应商的 SOM 约 $3 billion 是一个低置信度但说得通的规划启发式,直到管理层提供管线、胜率和合同金额数据之前,都不应把它视为硬结论。 [CM001, CM002, CM003, CM004, CM005, CM006]

市场规模表
市场 / 细分2024 / 当前规模2030 / 2035 预测CAGR来源 / 注意事项
AV 仿真~$2.5B到 2030-2035 年达 $7-8B~13-20%MarketsandMarkets、Grand View、Allied、Mordor 和 Business Research Insights 的共识;定义会因是否纳入数字孪生和验证工具而不同
自动驾驶软件2024 年未清晰披露到 2035 年约 $7B~13.3%MarketsandMarkets 预测;范围宽于纯仿真,可能与 OEM 内部软件支出重叠
ADAS~$33B到 2030 年超过 $80B低十几位数Mordor Intelligence 的 ADAS 市场视角;属于相邻类别,而非 Applied Intuition 的纯核心 TAM
国防自主系统约 $2B 可寻址工具视角到 2030 年估计 $4-5B中十几位数估计基于与自主系统相关的美国国防支出和项目活动推导;没有单一已发布市场报告单独拆出外部工具支出

市场估计来自第三方研究机构和公共支出视角;Applied Intuition 的具体份额、内部细分和已实现收入暴露未公开披露。

[CM001, CM002, CM003, CM004, CM034, CM035]
市场细分拆解表
细分可寻址市场估计Applied Intuition 位置渗透率备注
汽车 OEM 工具约 $6-7B 实际 SAM 视角核心市场公开匹配度最高;公司称在前 20 大 OEM 中深度渗透这与 Applied Intuition 当前仿真、验证和车辆软件栈最匹配
商用卡车~$2-3B强相邻领域有选择性但战略重要高速公路自动驾驶的 ODD 更窄、ROI 更清晰,但部署周期仍长
国防~$2-3B快速增长的扩张通道公开定位很强;具体合同基础不透明国防预算与消费者 AV 情绪的相关性可能更低,也会奖励验证密集型工作流
工业自主系统约 $1-2B 可选相邻机会新兴 / 期权价值公开渗透度低于汽车或国防采矿、农业和建筑在逻辑上有吸引力,但 Applied 相关的当前规模披露不足

可寻址细分估计是基于产品匹配、公开定位和第三方市场区间得出的方向性视角,而非公司披露的收入结构。

[CM005, CM006, CM007, CM008, CM009, CM038]
FM001: 按细分市场看市场规模

对 Applied Intuition 相关的主要自主系统周边赛道,横向比较 2024 年估算与远期端点市场规模。

AV 软件采用 2035 年端点,其他类别更接近 2030 年;图表用于比较相对规模,不是单一年份的预测截面。

[CM001, CM004]
FM002: TAM / SAM / SOM 漏斗

从广义自主系统周边支出,到外部工具供应商更窄的可服务市场,形成一层市场规模漏斗。

TAM、SAM、SOM 是基于多份市场报告和公开支出数据拼出的口径;没有一项是公司披露数字。

[CM034, CM035]

2.2 监管环境

监管环境之所以影响 Applied Intuition,不是因为它设置了单一联邦“放行 / 不放行”闸门,而是因为它在多个司法辖区持续制造验证负担。NHTSA 当前框架仍没有独立的联邦自动驾驶车辆上市前认证制度;市场通过既有安全标准、缺陷权力、常设命令和各州许可结构运转。这种碎片化具有商业意义。它会拖慢商业化,因为客户必须处理不一致的法律和运营要求;但它也强化了仿真和基于场景验证的必要性,因为 OEM 和开发者在部署前需要证明车辆覆盖大量运行条件。 在商用卡车上,面向 FMCSA 的规则清晰度仍不完整,即便高速公路自动驾驶仍是行业中最清楚的运行设计域之一,全国范围去驾驶员部署的时间也仍不确定。对全球 OEM 项目而言,EU 和 UNECE 路径又增加了一层型式认证工作和时间复杂度。防务则完全走另一条路:采购不是围绕一套民用 AV 规则书,而是由项目特定的安全、互操作和测试要求塑造。结果是,监管不只是逆风。它既是延迟因素,也是 Applied Intuition 所在品类的需求驱动器;每新增一层报告要求、安全命令或辖区差异,测试覆盖、场景管理和验证证据的工作量都会增加。 [CM010, CM011, CM012, CM013, CM014, CM015]

监管格局表
司法辖区机构当前状态时间线对 Applied Intuition 的影响
美国NHTSA指南、长期命令和现有安全权力仍是主要联邦框架;没有独立的上市前 AV 认证制度2026-2028 年通过渐进规则制定和报告要求演进高:增加验证和安全证据工作量,同时让部署节奏保持碎片化
美国FMCSA商用自动驾驶卡车规则仍不完整,并且在对照现有驾驶员要求时解释不一可能逐步澄清,而非一步到位审批高:放慢全国范围无驾驶员卡车时间线,但保留对场景测试和合规工作流的需求
EU / 全球 OEM 认证UNECE / WP.29框架继续按规则集和成员国实施路径演进2026-2029 年分阶段采用和解释中高:全球 OEM 客户必须按更多认证和文档路径完成验证
美国国防DoD / 项目标准采购仍按项目定制、由任务驱动且对安全敏感,而不是由一本民用 AV 规则书统管持续多年的项目节奏高:有利于灵活的仿真、测试和车辆软件基础设施,并按项目需求定制

监管状态汇总自 NHTSA 页面、Federal Register、国会材料和公开国防采购背景;具体到每个客户的合规负担会因车辆类别和部署领域而不同。

[CM010, CM011, CM012, CM013, CM014, CM015]
FM003: 监管就绪度 KPI

紧凑呈现市场当前监管形态:碎片化、合规负担重,且仍缺少全国统一的 AV 审批机制。

州级框架数量是基于政策摘要和行业跟踪得出的 2025-2026 年方向性估计;不是 NHTSA 官方指标。

[CM012, CM013]

2.3 市场动态与竞争格局

Applied Intuition 竞争的是一个碎片化工具市场,而不是边界清晰的软件类别。客户通常不会购买“一个 AV 平台”;他们会拼装仿真、数据管理、地图、感知开发、验证和安全论证工具的组合。这种碎片化给覆盖广泛工作流的基础设施供应商创造机会,但也意味着采购可以保持模块化,并对价格敏感。开源工具和内部工程栈会在堆栈低端施压,成熟买家也可能同时使用多家供应商。因此,Applied Intuition 的战略优势不在于垄断式市场份额,而在于跨汽车、卡车和防务解决最难的集成与验证工作流。 多个顺风因素支撑品类增长。OEM 数字化转型和软件定义车辆项目,会在实体原型完成前把更多验证工作前移到软件中。ADAS 强制要求和安全预期会扩大相邻支出,即便纯 robotaxi 时间线继续滑坡。传感器成本下降和数据量增长,会改善自动驾驶开发经济性;数据飞轮也让高保真仿真随时间变得更有价值。同时,市场逆风是真实存在的:漫长商业化周期拖慢收入确认,安全事故已经压低投资者热情,不同市场报告按范围和地域给出的结果差异很大。因此,这个市场有吸引力,但并不自动容易。品类需求具备结构性,但从技术相关性走到快速收入增长,仍卡在采购周期、监管清晰度和客户项目存活率上。 [CM017, CM018, CM020, CM021, CM022, CM023]

AV 市场增长驱动因素表
驱动因素强度时间线受益方证据
ADAS 强制要求与安全压力当前至 2030 年OEM 验证平台及相邻自主系统工具厂商ADAS 市场增长和安全预期抬升,会扩大验证支出;即便 Level 4 乘用车时间表延后,这条逻辑仍成立
OEM 数字化转型当前至多年期Applied Intuition 核心汽车工作流软件定义汽车项目把更多测试和验证前移到以仿真为主的开发循环
DoD 自主系统支出中高2026-2030面向国防的仿真与车辆软件栈公开支出数据和公司国防定位显示预算动能真实存在,但合同透明度有限
LiDAR 与传感器成本下降持续新兴自主系统项目和新进入者硬件和算力成本下降,扩大试验范围,也让仿真支撑的开发在经济上更可行
数据飞轮复利式累积拥有可复用场景和边缘案例库的规模化工具厂商里程、场景和项目反馈越多,仿真保真度和验证价值越高
开源竞争中等逆风持续存在低端工具买家和价格敏感团队开源和内部栈带来价格压力,也让采购更模块化;即便如此,企业买家仍会为集成和支持付费

“强度”是对 Applied Intuition 所在品类需求影响的方向性判断;最后一行是结构性约束,而不是单纯增长顺风。

[CM017, CM018, CM019, CM020, CM021, CM022]
FM004: 自主系统买方到部署流程

OEM、卡车和防务渠道的买方,如何把仿真与验证工具转化为部署就绪。

[CM023, CM038]

2.4 防务与政府板块

防务不是 Applied Intuition 可有可无的相邻业务;公开公司材料显示,它正被定位为公司的核心扩张向量之一。逻辑很有说服力。民用 AV 支出仍受商业化延迟和公共安全审视约束,但防务自动驾驶预算可以用任务效果、后勤韧性、训练和生存性来证明合理性。这会改变买方行为。政府项目以项目为单位、重视安全敏感性和验证,天然更贴近仿真、测试和车辆软件基础设施,而不是面向消费者的网约车经济学。公开联邦支出数据也显示,自动驾驶相关项目已代表每年数十亿美元义务支出,即便仅从搜索结果仍难把精确工具层支出单独剥离出来。 对 Applied Intuition 来说,防务在战略上做了两件事。第一,它把可触达市场扩到乘用车项目之外,降低公司对单一 AV 商业化时间线的依赖。第二,它提升了软件栈的定性价值,因为防务用户高度重视场景生成、数字测试和任务特定验证。限制在于透明度。确切项目金额、合同时间、以及自动驾驶预算中哪些部分真正可由 Applied Intuition 触达,都很难从公开来源锁定。因此,防务应被视为有意义的顺风和战略对冲;但没有管理层级别披露之前,它还不是一个已完全量化的收入桥。 [CM008, CM019, CM030, CM031, CM032, CM033]

2.5 市场风险与行业逆风

这个市场的主要风险不是自动驾驶软件是否重要,而是商业化能否按投资者时间线到来。工具供应商可以在客户规模化部署前很多年就赢得项目,这会拉长销售周期,并把收入兑现推到很远。RAND 长期以来的安全论点和更近期的 AV 覆盖,都强化了同一结论:证明安全既昂贵又缓慢,也不太可能只靠道路里程解决。这支撑仿真需求,但也意味着客户往往多年停留在评估或有限部署状态。安全事故和公开 AV 挫折,已经把市场叙事从普遍 robotaxi 热情转向 ADAS、高速卡车和防务等更窄领域。 另一个关键尽调风险是不透明。公开来源不披露 Applied Intuition 在仿真和自动驾驶工具中的确切收入、ARR 或市场份额,因此投资者无法把公司估值直接连接到某个已知品类支出占比。第三方市场估计也会因衡量对象不同而出现重大差异:只看仿真、看更宽的自动驾驶软件,还是纳入相邻 ADAS 和数字孪生预算,结果都不一样。结论是,这个市场显然重要且在增长,但仅凭公开证据仍难精确承销。投资者应把本章市场测算视为有边界的分析镜头,而不是替代私人管线、定价、胜率和分部收入数据。 [CM016, CM025, CM026, CM027, CM029, CM036]

2.6 展品

Chapter 03

03竞争对手

3.1 仿真与测试工具竞争者

Applied Intuition 最直接的竞争对手,仍是向汽车工程团队销售仿真、测试和验证工具的既有厂商与初创公司。dSPACE 是最重要的既有厂商,因为它通过 ASM、VEOS 等产品,已经贴近 HIL / SIL 预算和长期 OEM 工作流。ANSYS 带来另一类威胁:它不像 dSPACE 那样汽车原生,但 AVxcelerate 叠加广泛工程软件信用和企业采购触达。IPG Automotive 的 CarMaker 在车辆动力学、ADAS 和场景验证流程中尤其重要,这意味着买家可以继续拼装模块化工具链,而不必把 Applied 作为主平台供应商。Cognata 更多在合成场景和云仿真上竞争,Metamoto 和 VectorCAST 则代表仿真与嵌入式测试层中更窄但仍相关的替代方案。 实际含义是,Applied 不是在防守一个单一巨型对手。它面对的是一组点状工具:即使没有任何一家能覆盖 Applied 全栈的宽度,也可能在某个工作流里“够用”。这对采购很重要,因为汽车买家常常同时使用多套工具,而不是立即标准化到一家供应商。Applied 的优势因此在客户想要打通仿真、验证、数据和车辆软件开发时最强;如果客户只要某个狭窄技术任务的最佳单点工具,优势会弱一些。 [CP001, CP002, CP003, CP004, CP005, CP006]

竞争对手对比表
公司类型成立时间融资 / 收入核心产品差异化点Applied 相对定位
dSPACE直接竞争的既有仿真 / 测试厂商1988私营;具工业既有厂商规模,确切收入未公开ASM; VEOSHIL/SIL 积累深,卡位 OEM 工作流Applied 在仿真、数据、SDS 和国防上覆盖更宽
ANSYS / AVxcelerate直接竞争的既有仿真套件1970上市工程软件规模;有 Synopsys 收购背书AVxcelerate物理仿真覆盖广Applied 更汽车原生,也更全栈
Cognata直接竞争的创业公司同业2016风投支持;规模小于 AppliedOneSim; AVBox合成数据和场景生成Applied 拥有更大 OEM 触达和更宽软件范围
IPG Automotive / CarMaker直接竞争的既有仿真厂商1984私营工程软件厂商CarMakerOEM 测试中的车辆动力学可信度Applied 向数据、验证和车辆软件延伸更深
Scale AI相邻数据平台2016~$1.5B 总融资(估计)Scale 汽车数据平台标注、评测和数据运营Applied 在闭环仿真和车辆工具上更强
CARLA(开源)替代品 / 开源2017$0 许可成本CARLA免费、可扩展的 AV 仿真器,研究采用度高Applied 凭企业支持、集成和匹配 OEM 流程取胜
Wayve端到端 AV 平台2017~$2.8B 总融资(估计)具身 AI AV 栈具身 AI 叙事和 OEM 心智Applied 向多客户销售基础设施,而不是押注单一 AV 项目
Aurora端到端 AV 平台2017公开市场资本口径约 $1.0B+Aurora Driver聚焦卡车商业化Applied 覆盖更多客户,不依赖单一运营商模式
Waabi端到端 AV 平台2021融资规模估计约 $1.0BWaabi DriverAI 优先的虚拟 AV 开发Applied 拥有更广 OEM 准入和国防姿态
Metamoto聚焦型仿真创业公司2012私营;所审页面未公开融资Metamoto 仿真平台聚焦 AV 仿真工作流Applied 规模、OEM 触达和平台宽度更强

本表选取部分竞争对手,并非完整版图。若竞争对手公开页面没有披露确切数字,融资和收入单元格只作方向性判断;本表用于比较威胁形态,不是精确重建股权结构表。

[CP001, CP002, CP007, CP009, CP011, CP018]
功能对比矩阵
功能Applied IntuitiondSPACEANSYSCARLAScale AI
仿真保真度高;定位于闭环和全项目验证高;HIL/SIL 级传承高;工程焦点偏重物理中;研究基线强直接能力低;不是仿真优先产品
数据管理规模高;公司强调大规模数据工作流高;公司核心强项
国防认证公开国防定位强不是公开可见的核心差异化点所审页面未知 / 不突出公开强调有限
全栈软件是 — SDS + Vehicle OS + 工具否 — 点状工具否 — 仿真套件否 — 仅仿真器否 — 仅数据层
OEM 渗透高;宣称覆盖前 20 大 OEM 中的 17 家高;传统汽车客户基础深中;企业触达低;直接企业渗透中;通过汽车数据项目触达
定价企业级 / 报价制企业级 / 报价制企业级 / 报价制免费 OSS企业级 / 按用量
开源
云原生公开定位强所审页面显示不一 / 不清晰企业姿态不一开发者自管

单元格是基于官方产品页和文档综合得出的定性判断。公开证据较弱时,表述为不一、有限或不清晰,而不当作硬事实。

[CP009, CP011, CP012, CP013, CP014, CP015]
FP001: 竞争定位图——资本与规模视角

对 Applied Intuition 及选定直接、相邻和替代竞争者,做方向性的融资与规模对比。

数值是用于对比的合成规模估计,不是经审计的融资总额。dSPACE 和 ANSYS 充当既有厂商规模代理,而非风投式融资数字;CARLA 因开源,许可融资压力记为零。

[CP025, CP026]

3.2 平台与数据竞争者

Applied Intuition 还面临来自相邻平台的显著压力,这些平台控制数据、开发者工作流或低成本替代品,而不只是传统仿真软件。Scale AI 是最清楚的相邻竞争者,因为现代自动驾驶买家不会孤立地给仿真做预算;他们也会购买数据标注、整理、评估和模型反馈基础设施。拥有这些工作流的供应商可以向上游移动,进入验证和分析,即使起点并不是仿真器。因此,Scale AI 是一个新兴竞争面,而非传统正面同业。 CARLA 制造的是另一种压力。它不是 Applied OEM 级工作流的交钥匙企业替代品,但对研究团队、初创公司、大学和价格敏感试点而言,它是有效的零授权费基准。因此,CARLA 会压缩基础仿真能力的支付意愿,并给内部工程团队一个可信的“围绕开源自建”选项。Applied 在支持、集成、企业部署和流程适配上看起来仍占优;但 CARLA 的存在意味着,公司不能只靠软件稀缺性卖价。它必须把保真度、工作流集成和价值兑现速度,卖到免费或数据中心化替代方案之上。 [CP011, CP012, CP013, CP014, CP015, CP030]

FP003: 竞争威胁漏斗

威胁从最直接的点状工具竞争,收窄到没那么直接但仍重要的替代和平台风险。

分数是方向性威胁强度分,不代表市场份额。

[CP001, CP011, CP018, CP032]

3.3 端到端 AV 竞争者

Applied Intuition 面临的一些最重要竞争压力,来自根本不想销售点状工具的公司。Waymo、Mobileye、Wayve、Aurora 和 Waabi 会塑造客户对“胜出”自动驾驶架构应是什么样的预期。Waymo 最重要,因为它是技术可信度、数据集、仿真成熟度和人才吸引力的基准,尽管它目前不销售通用外部工具栈。Mobileye 重要,是因为它已经拥有深厚 OEM 关系和可授权的车辆智能姿态,在某些账户中可能降低客户对独立第三方工具供应商的需求。Wayve、Aurora 和 Waabi 重要,是因为它们围绕一个想法争夺 OEM 心智:买家也许更偏好端到端自动驾驶平台或战略伙伴,而不是模块化工具环境。 因此,这些公司是间接但具有战略意义的竞争对手。它们可以改变客户预算去向、技术路线图形态,以及仿真是作为独立类别采购,还是被捆进更宽的车辆软件决策中。Applied 的挑战,是证明基础设施宽度和集成速度,比押注任何一种端到端自动驾驶论点更有价值。 [CP018, CP019, CP020, CP021, CP022, CP023]

FP002: 市场定位流——Applied Intuition 与点状工具

Applied Intuition 覆盖的 AV 软件栈更广;点状工具竞争者仍主要集中在仿真。

该流程是概念图,基于产品范围证据合成,不是公司披露的架构图。

[CP009, CP010, CP024, CP035]

3.4 Applied Intuition 的竞争优势

Applied Intuition 的核心差异点在于,它并不把自己定位成单纯仿真供应商。公司材料把产品框定为更宽的 Physical AI 和车辆软件平台,横跨仿真、数据、验证、SDS、Vehicle OS 和防务工作流。这很重要,因为当买家想要一家供应商加速堆栈多层部署,而不是管理多个点状工具集成时,竞争护城河最强。公司还声称拥有异常深的客户渗透,服务全球前 20 大汽车 OEM 中的 17 家;如果方向上属实,这会带来多数小型挑战者很难复制的分发杠杆。 防务进一步强化差异化。公开材料显示 Applied 能支持涉密或安全敏感用例,这是开源工具和很多纯商用竞争对手不强调的姿态。最后,公司的上市速度叙事——常被概括为帮助客户用数周而非数年部署——把工作流压缩变成竞争论点。限制在于,宽度是一把双刃剑:Applied 卖的堆栈越宽,邀请进来的竞争者就越多,覆盖仿真、数据、操作系统基础设施和端到端自动驾驶平台。 [CP009, CP016, CP017, CP024, CP033, CP034]

FP004: 竞争护城河 KPI 快照

用紧凑数字视图呈现 Applied Intuition 护城河最关键的竞争面。

计数基于章节分类和公司说法合成,不是竞争者报告的 KPI。

[CP009, CP016, CP018, CP024, CP035]

3.5 竞争风险与被替代场景

主要替代风险不是某个对手一次性复制 Applied Intuition 的所有产品,而是客户判断自己不需要完整套件。dSPACE 或 ANSYS 可以在验证密集型项目中继续盘踞;CARLA 可以给早期团队的价格封顶;Scale AI 可以拿下数据与评估层;Mobileye、Wayve、Aurora 或 Waabi 可以说服 OEM:更宽的平台关系比采用模块化工具更具战略价值。公开证据也没有显示 Applied 拥有排他性 OEM 合同或持久单一来源地位,因此多供应商并用风险看起来真实存在。在大型车辆项目中,买家通常会同时保留不止一条仿真、数据或验证路径。 另一个承销问题是不透明。竞争对手的准确收入、已实现 ACV 和 OEM 账户份额大多不公开,因此威胁排序本来就是近似的。这意味着关键尽调问题不是 Applied 是否抽象地具有差异化——它显然有——而是这种差异化能否撑过续约周期、在账户内拓宽,并防止采购重新碎片化,流向既有厂商、开源或端到端平台替代方案。因此,投资者应索取胜 / 负数据、替换 dSPACE 级既有厂商的证据,以及合同细节,以判断 Applied 正在成为系统记录层,还是只是更大堆栈中的一个工具。 [CP025, CP026, CP027, CP028, CP029, CP030]

竞争威胁评估
竞争对手威胁等级原因相对 Applied 的护城河尽调问题
dSPACE中高深嵌验证台架和既有 OEM 测试工作流Applied 的护城河在于栈更宽,但既有厂商卡位风险真实存在拿出 Applied 替换或挤掉 dSPACE 这类工具的赢单 / 输单数据
ANSYS大型企业软件基础和仿真可信度Applied 看起来更汽车原生,也更偏部署梳理具名 OEM 账户和 RFP 的重叠
Cognata低-中在合成数据和云仿真上有可信细分位置Applied 分发更强、栈更宽量化合成场景采购中的正面交锋结果
CARLA低-中(价格)面向研究团队和试点的免费开源替代方案Applied 的护城河是企业支持、工作流集成和服务质量衡量 CARLA 或内部栈多常压低定价或拖慢转化
Scale AI中等新兴控制标注、评测和数据运营后,可向上游扩张Applied 今天在闭环仿真和车辆软件集成上更强识别共同账户,并判断数据买家之后是否会向同一供应商购买仿真
Aurora直接威胁低 / 端到端威胁高不卖点状工具,但争夺卡车预算和 OEM 架构决策Applied 的护城河是多客户工具,而不是单一运营商平台测试 OEM 在卡车和自主系统项目中是否更偏好平台合作而非模块化工具

威胁等级反映对 Applied Intuition 赢单、扩张或留住账户能力的潜在影响,而不是简单看公司规模。最高风险不是单一“击倒型”竞争对手,而是预算被既有厂商、开源和端到端平台切碎。

[CP027, CP028, CP029, CP030, CP031, CP032]

3.6 展品

Chapter 04

04财务

4.1 融资历史与资本结构

Applied Intuition 的融资历史,更适合理解为一连串越来越强的验证点,而不是一组披露完整、干净的轮次。公开证据支持:2017 年由 Andreessen Horowitz 支持的一笔小额 Series A;2018 年由 General Catalyst 领投的约 $40 million Series B;2019 年 $125 million Series C,使披露融资总额约达 $175 million;2020 年估值 $3.6 billion 的 $175 million Series D;2024 年估值 $6 billion 的 $250 million Series E;以及 2025 年按 $15 billion 估值定价的 Series F。关键限制在于披露质量。Series A、Series B 和 Series F 的确切现金到账并未完全公开,因此累计融资额应仍被视为约 $1.5 billion 或以上的估计,而不是硬事实。 这种模糊性会影响承销,因为估值信号远比现金流信号清楚。SEC EDGAR Form D 搜索结果显示,Applied Intuition 至少在部分融资中使用了私募备案,但这些通知不能替代经审计财务报表、详细股权结构披露或当前现金余额。因此,公司呈现出一个看似资金极其充足、却仍难以建模稀释、剩余流动性和所有权集中度的资本结构。 [CI001, CI002, CI003, CI004, CI005, CI006]

融资轮次表
轮次日期金额($M)估值($B)领投方累计融资备注
Series A 轮2017约 2(估计);未披露Andreessen Horowitz约 2(估计)早期种子 / Series A 规模来自后续轮次回顾推算;确切现金到账和估值未公开
Series B 轮2018约 40(估计);未披露General Catalyst约 42(估计)公司和市场数据回顾均提到 General Catalyst;公开轮次条款仍有限
Series C 轮2019125投资方包括 Lux Capital、a16z、General Catalyst~175官方新闻稿披露 $125M,并称累计融资约 $175M
Series D 轮20201753.6投资方包括 Elad Gil、Addition、Coatue~350Series D 也加入 Rick Wagoner 和 Dieter Zetsche 担任顾问
Series E 轮2024-102506.0投资方包括 Lux Capital、Porsche Investments、Elad Gil~600官方文章称公司已经盈利,并保持可持续的三位数同比增长
Series F 轮2025-10约 500(估计);未披露15.0投资方包括 BlackRock、Kleiner Perkins、Fidelity、Lux Capital约 1,500+(估计)估值和投资方名单证据充分;确切轮次现金金额公司未披露

融资轮次整理自公司新闻稿、市场数据档案,以及截至 2026 年 5 月的 2025 年融资报道;Series A/B/F 的确切金额并未完全公开,因此累计融资仍为估计值。

[CI001, CI002, CI003, CI004, CI005, CI006]
FI001: 融资轨迹——估算轮次规模视角

用一致的轮次规模口径覆盖六轮融资,因为公开估值披露只从后期轮次开始。

由于公开估值数据只有从 Series D 起才清晰披露,所有轮次都用轮次规模而非估值。Series A、Series B 和 Series F 柱包含估计值。

[CI001, CI002, CI003, CI004, CI005, CI006]
FI004: 融资年表与估值拐点

从创立到最新 $15B 估值信号的近似融资年表。

早期轮次和 2025 年日期按月取整,因为审阅的公开来源披露年份或月份比准确日期更清楚。

[CI001, CI002, CI003, CI004, CI005, CI006]

4.2 关键投资者与战略意义

投资者名单现在比原始融资金额更具战略意义。Lux Capital 看起来是跨多个轮次最持久的支持者之一,Andreessen Horowitz 和 General Catalyst 则支撑了早期机构融资叙事。2024 年轮次加入 Porsche Investments,重要性在于它更容易被解读为战略汽车信任票,而不是纯财务支票。2025 年轮次意义更大:BlackRock、Kleiner Perkins、Fidelity 和 Lux 一起出现,说明 Applied Intuition 不再只是按专业自动驾驶工具初创公司的方式融资。它正被作为更宽的基础设施资产融资,对跨界投资者和机构资金都有吸引力。 与最新融资同期宣布的 OpenAI 合作,进一步强化了这种解读。它暗示投资者正在承销一个可能从仿真延伸到更宽 Physical AI 软件层的平台,覆盖汽车、卡车、防务系统和其他机器。这并没有消除尽调问题;它增加了问题。投资者仍需要所有权比例、董事会或观察员权利、按比例跟投条款、二级出售历史,以及最新资本提供方的预期退出期限。 [CI011, CI012, CI013, CI014, CI034, CI037]

投资者画像表
投资者类型轮次战略意义治理权利尽调问题
BlackRock跨界 / 机构资产管理公司Series F 轮显示主流机构对物理 AI 基础设施有需求未公开披露确认持股比例、董事 / 观察员权利和后续跟投意愿
Kleiner Perkins风险投资Series F 轮增加顶级风投品牌,并带来企业 / 深科技网络触达未公开披露确认持股规模、按比例认购权和退出预期
Lux Capital风险投资Series C、Series E、Series F 轮参投自主系统和国防相邻平台的长期多轮支持者未公开披露梳理各轮持股及任何特殊信息权
Andreessen Horowitz风险投资Series A、Series C 轮参投最早的机构背书和网络杠杆未公开披露确认持股仍主要来自新股、已被稀释,还是部分来自老股交易
General Catalyst风险投资Series B、Series C 轮参与方帮助公司越过最早期阶段并继续放大未公开披露确认当前持股以及是否仍有董事会影响力
Elad Gil成长轮投资人Series D、Series E 轮参与方在后期扩张中释放运营者型投资人背书未公开披露厘清持仓规模和融资影响力
Addition / Lee Fixel成长轮投资人Series D引入偏创始人友好的成长资本取向未公开披露确认持股比例及是否有老股交易历史
Coatue跨市场成长轮投资人Series D在首次大幅估值跃升时引入公私市场视角未公开披露确认是否仍持有该仓位及具体条款
BOND / Mary Meeker成长轮投资人Series E支撑成长期叙事,并向更广市场释放信号未公开披露厘清信息权和预期持有期限
Porsche Investments汽车战略投资人Series E增加贴近 OEM 的战略可信度未公开披露确认是否存在商业、平台或排他性权利
OpenAI战略合作方Series F 轮时期合作将叙事从自主系统工具扩展到物理 AI 基础设施合作条款未公开要求提供收入分成、联合 GTM 或排他性细节

具名投资人来自公开融资公告和媒体报道。治理安排、董事会权利和持股比例仍是私营公司尽调事项,不是公开事实。

[CI011, CI012, CI013, CI014, CI034, CI037]
FI002: 投资人结构成熟化漏斗

随着故事线转向 Physical AI,Applied Intuition 的股权结构表从早期 VC 扩展到成长、战略和跨界资本。

由于持股比例和治理条款未公开,数值统计的是投资人类别信号,而非所有权比例。

[CI011, CI012, CI013, CI014, CI034, CI037]

4.3 财务表现与披露指标

对一家估值 $15 billion 的公司而言,Applied Intuition 披露的常规经营数据很少。公开材料中最强的正面事实,是管理层在 Series E 文章中称公司已经盈利,并以可持续的三位数百分比同比增速增长。如果准确,这一组合对于资本密集型自动驾驶相邻软件公司而言并不常见,也有助于解释为什么投资者愿意在下一轮以高得多的价格为业务融资。公开规模代理指标也存在:公司称拥有 1,000+ 名工程师,并服务全球前 20 大汽车 OEM 中的 17 家。 但这些信号没有解决核心承销问题。收入、ARR、毛利率、现金余额、客户集中度、NRR 和烧钱率,在已审阅公开来源中均缺席。PitchBook 和 CB Insights 有助于框定估值和融资历史,却无法补上经审计收入质量或利润率路径的空白。因此,Applied Intuition 从叙事和投资者质量看像一家强大的私营软件基础设施公司,但在支撑详细财务模型所需的确切指标上,仍异常不透明。 [CI015, CI016, CI017, CI018, CI019, CI020]

财务 KPI 快照表
指标数值置信度日期来源缺口
估值$15B2025官方 Series F 文章以及 Bloomberg/CNBC 报道该轮确切现金融资额仍未披露
累计融资~$1.5B+2026-05-19根据已披露和估计轮次测算Series A/B/F 金额未完全披露
收入2026-05-19所审官方和市场数据来源没有公开数值需要经审计收入历史和分业务线结构
ARR2026-05-19所审官方和市场数据来源没有公开数值需要 ARR 桥、NRR 和队列瀑布
毛利率2026-05-19所审官方和市场数据来源没有公开数值需要分业务线毛利率桥和服务负担
员工数1,000+ 名工程师2026-05-19公司介绍页和招聘材料需要按研发、销售、支持和国防拆分总 FTE
客户数全球前 20 大汽车 OEM 中的 17 家2025-2026公司材料和 Series F 叙事渗透率口径没有揭示 ACV、集中度或留存
盈利状态已盈利;可持续的三位数同比增长2024-10Series E 官方文章需要经审计盈利口径、利润率定义和现金流转化

置信度反映证据质量,而不是业务质量。null 值表示尽管公开融资叙事丰富,这些指标仍未公开。

[CI005, CI006, CI008, CI015, CI016, CI017]
FI003: 财务披露画像 KPI

紧凑呈现市场看得见的事实,以及仍未公开的指标。

该 KPI 视图有意统计披露覆盖度,而不是声称公开来源无法证明的收入质量。

[CI015, CI017, CI018, CI019, CI027, CI028]

4.4 Embark 收购与并购策略

Embark 收购在财务上重要,原因有两点。第一,它是一笔困境资产购买,而不是传统高溢价增长型收购。Applied Intuition 宣布交易价值约 $71 million,以股票支付,并随后在 2023 年 8 月 2 日宣布完成。公开材料把这笔交易描述为在行业失败之后获取有用的卡车软件、数据和人员,而不是为了大规模购买收入。这很重要,因为它意味着在许多自动驾驶公司难以为商业化融资的时点,Applied 可以不用投入大量现金就扩展产品范围。 第二,Embark 也是本章最清晰的负面信号。Embark 的代理与投资者材料显示,即便是一家公开上市、由风险资本支持的 AV 卡车公司,也无法撑过商业化缺口。对 Applied Intuition 而言,正面解读是它低价买到了战略资产。负面解读是,周边市场仍很残酷,Applied 自身估值也必须放在一个已经出现知名资本毁灭案例的行业中审视。 [CI022, CI023, CI024, CI025, CI026, CI032]

4.5 财务风险画像与资本效率

主要财务风险不是明显资不抵债,而是估值与披露不匹配。Applied Intuition 的公开估值从 2020 年 $3.6 billion 升至 2024 年 $6 billion,再到 2025 年 $15 billion,但公开 KPI 披露并未同步扩张。最新融资在估值和投资者名称上更容易核实,在确切现金金额或当前经营结果上则难得多。这不代表公司弱;它代表外部投资者被要求主要根据投资者质量、伙伴证明、盈利表述和规模代理指标来承销一个溢价,而不是根据完整经营数据包。 因此,资本效率比较只能近似处理。按披露风格看,Applied 比 Aurora 这类公司更像软件公司;但由于累计融资估计约 $1.5 billion 或以上,它又比许多纯工具供应商资本化程度更高。正确结论不是公司必然被高估,而是即便公开叙事很强,承销仍需要对 Series F 确切现金到账、客户集中度、NRR、董事会权利、稀释、现金余额和下一轮触发因素开展私人尽调。 [CI027, CI028, CI029, CI030, CI031, CI038]

资本结构风险表
风险严重程度描述缓释信号
私营公司不透明度审计财务、股权结构表和当前董事会构成均未公开公司多次拿到顶级资本,说明私下数据室很可能存在
收入 / ARR / 毛利率未披露支撑估值所需的核心运营 KPI 在公开来源中缺失盈利和三位数增长表述方向正面,但信息不完整
估值节奏在更充分 KPI 披露前,估值约一年内从 $6B 升至 $15B投资人质量和盈利表述部分解释了出价意愿
创始人稀释和治理未知中高公开来源未披露持股 %、董事席位或观察员权利Form D 证据确认私募发行流程,但不确认轮后持股
老股交易未知没有公开证据说明后续轮次是否包含有意义的老股流动性可能无害,但会影响激励一致性和价格发现
客户集中度和 NRR 未知中高合作伙伴标识和 OEM 渗透表述没有披露续约质量或头部客户依赖商业合作伙伴证据说明存在真实需求,但不能证明收入耐久性

本表聚焦公开来源尚未解开的财务结构风险,不涵盖其他章节已讨论的产品或市场风险。

[CI027, CI028, CI029, CI030, CI031, CI036]
可比公司资本效率表
公司累计融资($M)估值($B)倍数阶段备注
Applied Intuition~1,500+(估计)15n.m.盈利的私营平台公开视角只能看估值和盈利表述;收入仍未披露
Waymon/aAlphabet 支持的自动驾驶平台可作为规模上限参考,但母公司资金掩盖独立资本效率,不能直接按风险投资口径比较
Auroran.m.上市 AV 卡车公司可作为商业化成本基准,但当前市场估值波动大,不能直接与私营软件平台比较
Waabi~1,000(估计)n.m.AI 优先的私营 AV 公司公开 KPI 披露有限,同行融资规模只能作方向性参考
Scale AI~1,500(估计)n.m.私营 AI 基础设施公司相比全栈 AV 运营商,它更接近软件基础设施类比对象,但公开 KPI 披露仍不完整

同行行仅作方向性比较启发,不是经审计的一一对应可比公司。多家私营同行披露数据不足,无法计算干净的收入或利润率倍数;为避免虚假精确,相关位置使用 null 或 n.m.。

[CI038, CI039]

4.6 展品

Chapter 05

05产品与技术

5.1 产品组合概览

Applied Intuition 现在把产品呈现为面向所有移动机器的统一平台,而不是狭窄的自动驾驶工具供应商。公开表述反复把 Self-Driving System(SDS)、Vehicle OS 和 Tools for Vehicle Intelligence 放在更宽的 Physical AI 伞下,并将同一套堆栈专门化到汽车、卡车、采矿、农业、建筑和防务。这很重要,因为买家不是被要求授权单个仿真器;他们被要求采用一套集成软件底座,它可以从 ADAS 走向自动驾驶,从道路走向非道路,从商用走向防务。第二个值得注意的变化是品牌纪律。当前产品页面不再突出 Simian、Spectral 等旧名称,但底层仿真 / 评估和数据 / 质量控制功能,在 Tools for Vehicle Intelligence 叙事中仍清晰可见。从尽调角度看,产品组合宽、模块化且多领域,但各个表面的成熟度并不一致,部分最新层级读起来仍更像快速推进的平台叙事,而不是已被独立基准验证的产品证明。 [CE001, CE002, CE005, CE006, CE007, CE008]

产品组合表
产品 / 模块功能核心能力目标用户成熟度相比开源的差异化
SDS完整 ADAS / 自动驾驶软件栈覆盖陆、空、海的专用感知、计算和控制,并支持多载具编队OEM 自动驾驶团队、车队运营商、国防项目已公开商业化露出;成熟度因领域而异不止仿真,还延伸到可部署的自主系统和运营工作流
Vehicle OS机器操作系统层统一平台覆盖感知、规划、控制、可观测性和代码优先开发OEM 软件 / 平台团队和机器软件团队最近推出,但公司称其模块化且可用于生产提供可复用的 OS 和工作流底座,而不只是仿真器
Simian(仿真;旧品牌)场景仿真与评估当前公开页面显示,仿真和评估已纳入 Tools for Vehicle Intelligence验证工程师和自动驾驶开发团队旧名称不再突出,但功能清晰存在与数据和部署工具集成,而不是作为研究型仿真器单独存在
Spectral(数据管理;旧品牌)数据摄取、整理、标注和血缘PB 级摄取,以及服务训练、仿真和评估的闭环数据飞轮ML 平台和数据工程团队旧名称不再突出,但功能清晰存在将数据运营与仿真、评估和部署反馈回路配对
物理 AI 平台总平台叙事跨领域整合基础模型、仿真、自主系统、语音、生成式 AI 和开发者工具CTO、平台采购方、战略 OEM / 国防利益相关方当前旗舰定位将 Applied 定位为多领域基础设施,而不是单一自动驾驶 SKU
MCP 接口智能体编排层令牌化 UI 和可接入 MCP 的接口,支持端到端智能体驱动任务内部开发者、验证工程师、工作流负责人非常新,公开基准仍很少将工作流自动化推进到仪表盘和手工工具交接之外
国防 / 保密环境任务自主和对抗环境技术栈自主软件、仿真基础设施、任务系统,以及基于网状网络的协同控制国防项目办公室和军民两用主承包商公开已有实战式证明;保密细节仍未公开国防姿态和任务环境表述超过多数开源或纯商业工具

当前公开页面不再突出 Simian / Spectral 旧名称;上述行把用户提到的旧标签映射到现在归入 Tools for Vehicle Intelligence 的仿真 / 评估和数据管理功能。成熟度标签反映公开证据质量,而不是内部产品阶段访问权限。

[CE001, CE005, CE006, CE007, CE008, CE016]
FE001: Applied Intuition 产品架构

公开披露的架构把 SDS、Vehicle OS 和 Tools for Vehicle Intelligence 叠成一个 Physical AI 技术栈。

[CE001, CE008, CE011, CE013, CE015, CE016]
FE003: 按细分领域看产品成熟度

基于公开证明的具体程度和部署证据,对 Applied Intuition 主要产品线做方向性成熟度读数。

分数是从公开表面读出的方向性成熟度,不是内部发布闸门。数值越高,说明页面提供的具体部署和工作流细节越多。

[CE005, CE008, CE013, CE016, CE021, CE024]

5.2 仿真与数据管理平台

最清晰的技术重心,仍是仿真、验证和数据运营的组合。Applied 的 Physical AI 和自动驾驶页面描述了 PB 级摄取、整理、标注、评估,以及真实世界传感器数据的闭环复用;研究页面则把这个故事延伸到神经仿真、合成数据和大规模 ML 基础设施。落到实践中,这像是一种企业级尝试:把开放研究栈常常分开的数据采集、整理、模型训练、仿真、评估和生产反馈连接起来。与 CARLA 的比较在这里最有用。CARLA 是一个真正的开源仿真器,具备 Python API、场景工具、ROS 桥接和 OpenDrive 支持,但本质上仍是一个仿真器,客户必须把它装配进更宽工作流。Applied 的公开卖点是:工作流本身就是护城河。限制在于,公开页面仍缺少投资者最想看到的具体吞吐指标:每小时场景数、标注吞吐量、合成到真实迁移统计,或相对既有验证套件的硬基准差异。 [CE004, CE013, CE014, CE015, CE018, CE026]

技术能力矩阵
能力Applied IntuitionCARLA(开源)dSPACEANSYSScale AI是否差异化?
仿真保真度高:物理准确、闭环、多领域仿真叙事中高:强研究型仿真器,传感器和环境可配置高:覆盖创新链的验证既有厂商高:AVxcelerate 传感器准确、闭环,聚焦 SiL / HiL低:未定位为仿真套件广度上是,但开放性上不是
数据规模高:PB 级摄取、整理和闭环数据飞轮中:具备传感器 / 数据检索,但数据运营不是核心产品中:偏验证,但公开叙事不是 PB 级数据平台中:具备可追溯安全数据和传感器仿真,但数据平台中心性较弱高:明确定位为 AI 数据平台混合;单看数据平台聚焦度,Scale 更强
安全认证支持中:公司声称支持安全和法规合规,但公开点名的标准很少中低:有研究和 RSS / OpenDrive 集成,但企业认证负担仍在用户侧高:验证既有厂商地位意味着流程可信度高:安全论证、认证/准入和可追溯数据表述明确低:公开定位不是安全案例平台Applied 没有明确公开领先
国防 / 保密高:对抗环境和军民两用定位居核心位置低:开源研究基线,没有国防产品叙事中低:不是明显公开差异点中低:工程栈强,但营销不围绕对抗环境低:没有可见的公开国防定位
OEM 集成高:面向 OEM 项目的 OS + 自主系统 + 工作流栈中:灵活,但用户集成负担重高:长期占据汽车工作流既有位置高:以开放架构面向 OEM 和 Tier 厂商设计中:邻近数据工作流,而不是完整车辆栈全栈集成上是
云原生高:基于 SDK 的工作流覆盖云、本地部署和隔离网部署中:可分布式使用,但由用户自行管理中:公有云姿态不是主叙事高:明确云原生,并采用开放架构高:AI 平台定位以云优先部分成立
开发者 API中高:代码优先、SDK 和可编程工作流表述明确,但公开文档很少高:Python API 和文档化生态公开中:公开主页对开发者界面着墨较少中高:开放架构和仿真器互操作表述明确中:平台定位清晰,但本次审阅的公开技术文档有限相比 CARLA 不成立
智能体 AI高:可接入 MCP、智能体驱动工作流表述明确低:审阅页面没有可比的智能体产品层低:不是可见公开主题中低:仿真先进,但没有智能体工作流叙事中:有 AI 平台叙事,但不是车辆专用智能体工具

单元格为基于已审阅公开来源综合得出的定性判断,不是实验室基准。矩阵意在比较已披露的技术形态和工作流广度,而不是声称经审计的功能对等。

[CE025, CE026, CE027, CE028, CE029, CE031]
FE002: 技术栈对比

围绕 Applied Intuition 公开产品版图中的主要工具类别,给出综合广度分。

分数是从已披露仿真广度、数据操作、OS/控制栈覆盖、防务适配和开发者触点合成的 1-5 定性分;不是基准测试分数,也不是客户 NPS 数据。

[CE025, CE027, CE037, CE043]

5.3 Physical AI 与智能体开发

故事中最新的一层,是从车辆智能工具转向更宽的 Physical AI 平台。当前公开页面明确表示,这套堆栈要在云端、本地和气隙环境中编排自动驾驶工作流,也声称可通过标记化 UI 设计和支持 MCP 的接口实现智能体驱动操作。这不只是泛泛的品牌语言。研究页面讨论世界-动作模型、强化学习后训练、4D 重建和闭环仿真;Series F 文章则把这些能力连接到开发者工具、AI 智能体,以及面向汽车、防务和非道路领域的具身 AI 部署。由此形成一个说得通的产品论点:仿真和数据基础设施成为智能体式开发循环运行的操作环境。但这一层仍早。公开材料尚未给出具名 SDK 包、已发布 REST 参考文档,或智能体层的具体基准证据。本轮调研中,AWS Marketplace 路径甚至只观察到一个损坏的公开页面,这进一步说明开发者生态方向上可信,但透明文档还不足以支持外部无摩擦评估。 [CE014, CE016, CE017, CE018, CE032, CE034]

开发者生态集成表
集成项类型状态客户收益
AWS Marketplace云分发本次看到公开产品路径,但上架详情页返回 400如果上架有效,可简化云采购和试用流程
GitHub 开源开发者基准已审阅来源未显示 Applied 自有代码库,但 CARLA 提供了买家可比较的参考开源基线帮助客户量化企业集成 / 支持相对自建拼装的价值
MCP Protocol工作流编排Physical AI 页面明确声称支持支持横跨仿真、数据和验证任务的智能体驱动自动化
AUTOSAROEM 软件标准已审阅的 Applied 页面未公开验证如果支持,将显著降低 OEM 软件组织的集成摩擦
HIL systems验证 / 测试台架集成虚拟化测试表述明确,但未在公开材料中看到具名 HIL 合作伙伴或接口对量产项目中的 ECU、台架和安全案例验证很重要
CI/CD 流水线开发者工作流代码优先工作流、Python 建模和基于拉取请求的变更管理表述明确让车辆开发对齐标准软件工程的发布和评审流程

本表混合了已披露集成和重要尽调问题。围绕 SDK、代码优先工作流和可接入 MCP 的编排,公开叙事可信;但具体 Marketplace 元数据、AUTOSAR 支持和具名 HIL 集成仍文档不足。

[CE010, CE014, CE016, CE032, CE034, CE039]
FE004: 客户部署漏斗

从原始车队数据到已部署自主系统或任务软件的概念性客户路径。

数值是示意性工作流强度分,不是转化率;目的是展示流程形态,而非账户数量。

[CE004, CE012, CE015, CE022, CE032, CE037]

5.4 防务技术平台

防务并不是副业。Applied 的防务页面称,公司为争议环境构建自动驾驶软件、仿真基础设施和任务系统,支持 mesh 网络上的协同自动驾驶,以及跨域多机器协同。最强的公开证明是 Army ISV 故事:Applied 称其用 10 天让 Infantry Squad Vehicle 实现自主化,将其与 Humvee 移动指挥所配对,并用 Vehicle OS 加越野自动驾驶软件,在实地作业中支持路线定义、进度监控和危险警报。这对产品尽调很重要,因为它说明公司可以把同一套软件底座适配到比乘用车更严苛、更不标准化的运行环境。承销限制同样重要。公开来源不披露这些环境中的涉密部分、确切认证边界或具名安全架构。因此,防务强化了差异化叙事,但尚未提供完整公开记录,足以在没有管理层访问的情况下承销任务关键场景所需的信任、认证和部署证据。 [CE019, CE020, CE021, CE022, CE023, CE036]

5.5 技术架构与开发者生态

公开证据最能支撑的架构叙事,是代码优先、模块化、面向集成。Vehicle OS 被描述为一层软件,横跨感知、规划、控制、车载软件、车外服务和云端工具;工程师用 Python 建模行为,用 pull request 管理变更。相比许多汽车买家仍然联想到的旧式、GUI 很重的验证链,这套形象更现代,也更软件原生。与此同时,竞争对手在一些领域披露得更细。CARLA 写明了 Python API、ROS bridge 和 OpenDrive 支持;Ansys 明确营销 SiL/HiL、ASAM 对齐、认证和传感器级准确仿真;CarMaker 披露 MIL/SIL/HIL/VIL、支持标准和 UN/ECE 审批流程;Mobileye 公开点名 RSS 和 ISO 9001。Applied 的公开页面则更强调监管标准合规、虚拟化测试、代码优先带来的速度和广泛互操作等结果。这足以证明它的技术栈有差异化,但还不足以核验 AUTOSAR 覆盖、具名 ISO 工作流、精确 HIL 集成,或基准级认证支持。若要把这套架构叙事视为已充分去风险化,上述问题仍是核心尽调项。 [CE010, CE011, CE012, CE027, CE028, CE029]

技术标准合规性
标准适用性Applied 覆盖情况竞争对手覆盖情况尽调需确认
ISO 26262汽车软件和工具资质认证的功能安全骨干公开页面称符合监管标准并支持安全验证,但本次未看到点名 ISO 26262 覆盖Ansys 和 CarMaker 更明确地营销安全论证、认证/准入和审批工作流要求提供 ASIL 分配、工具链资质、安全案例材料和评估方报告
ISO 21448 (SOTIF)对验证感知密集型 ADAS / AV 系统的安全行为很关键已审阅的 Applied 产品页面未点名竞争对手披露的安全验证工作流更明确,尽管已审阅页面也不总是点名 SOTIF要求提供场景覆盖方法、边界案例闭环指标和 SOTIF 工作产物
MISRA C++与量产车辆软件质量和静态分析纪律相关未看到公开 MISRA 声明已审阅的竞争对手页面也没有把 MISRA 放到主页级主张要求提供编码标准政策、静态分析栈和例外处理流程
AUTOSAR Adaptive对 OEM 集成、中间件复用和车辆软件可移植性重要公开的代码优先 OS 信息暗示有集成意图,但本次未看到点名 AUTOSAR 支持在已审阅集合中,竞争对手页面更强调开放架构和工具链集成,而非 AUTOSAR 具体声明要求提供具名 AUTOSAR 接口、支持版本和参考集成
ASAM OpenSCENARIO仿真可移植性常用场景交换标准未看到 Applied 点名声明Ansys 称 AVxcelerate 遵循 ASAM 标准;CarMaker 强调其支持的标准和接口要求提供导入/导出支持、场景库兼容性和客户证据
ASAM OpenDRIVE道路网络互操作性关系到仿真器和地图可移植性未看到 Applied 点名声明CARLA 文档明确提到 OpenDrive 支持要求提供 OpenDRIVE 支持细节和转换限制
SAE J3016定义客户和监管讨论中使用的 ADAS / AV 自动化等级Applied 公开将汽车 SDS 描述为从 L2+ 走向 L4 的路径Mobileye 和行业同行也使用明确的 ADAS 到 AV 等级框架要求按等级提供精确功能边界、后备假设和客户部署状态

本表区分点名标准的明确证据和更宽泛的安全 / 合规表述。Applied 的公开材料支持其严肃安全姿态,但不能充当认证登记表,因此多行仍是尽调需确认项,而非已验证合规结论。

[CE028, CE029, CE030, CE033, CE038]
FE005: 技术路线图时间线

公开可见的演进路径:从自主系统基础设施起步,到模块化 Vehicle OS、多领域案例研究,再到 2026 年 Physical AI 研究。

这是公开资料中的里程碑时间线,不是完整内部发布日志。日期来自当前公开文章或可见案例研究索引(如有)。

[CE044, CE045, CE046, CE047]
Chapter 06

06客户情况

6.1 客户概览与 OEM 渗透

Applied Intuition 的公开客户叙事,强在覆盖面,弱在具名账户深度。公司仍持续对投资人和客户强调,它服务全球前 20 大汽车 OEM 中的 17 家;公开页面也点名 Toyota、Volkswagen Group、General Motors 和 Hyundai 等汽车客户。这很重要,因为它意味着公司已经触达了全球乘用车买方群体中,对仿真、验证和自动驾驶工具最关键的大多数。它也说明 Applied 已足够嵌入汽车软件工作流,不只是在某个区域工程市场赢得关注,而是在北美、欧洲和亚洲都拿到心智份额。 尽调难点在于,覆盖面不等于深度。公开来源没有说明 17/20 说法中剩下的 OEM 是谁,也没有标明这些关系究竟是试点项目还是生产级工具链,更没有披露哪些账户贡献了有意义的经常性软件收入。因此,客户叙事显然真实,但只露出一部分。公开客户标识证据支撑了强市场渗透;但要有把握判断客户质量,还需要看到 ACV、续约韧性和生产绑定程度。 [CU001, CU002, CU003, CU004, CU005, CU006]

具名客户佐证表
客户细分领域合作关系类型证据合同深度(估计)成为客户时间
Toyota全球汽车 OEM合作伙伴 / 客户Toyota 合作伙伴文章;Applied 客户页面中低:公开关系已确认,生产部署深度未披露Pre-2026
Volkswagen Group全球汽车 OEM合作伙伴 / 客户Volkswagen 合作伙伴文章;Applied 客户页面中低:公开关系已确认,生产部署深度未披露Pre-2026
General Motors全球汽车 OEM合作伙伴 / 客户GM 合作伙伴文章;Applied 客户页面中低:公开关系已确认,生产部署深度未披露Pre-2026
Hyundai全球汽车 OEM合作伙伴 / 客户Hyundai 合作伙伴文章;Applied 客户页面中低:公开关系已确认,生产部署深度未披露Pre-2026
PACCAR商用卡车 OEM客户PACCAR IR 公告及 Applied 卡车业务文章高:客户侧宣布选择其支持自动驾驶卡车开发2020
Volvo Group商用车 OEM合作伙伴 / 客户Volvo Group 公告及 Applied 卡车业务页面中:有客户侧证据,合同范围未披露2022
U.S. Army国防 / 政府客户 / 项目用户Army.mil 文章;Applied 陆军相关文章;国防引用高:政府侧验证及在测引用Pre-2026
U.S. Air Force国防 / 政府客户 / 项目用户Applied 国防页面;Series E 轮文章中:具名引用,项目经济性未披露Pre-2026
NuroAV / 出行初创公司案例客户 / 历史引用Applied 案例页面;已失效的 nuro-customer URL中低:有历史证据,当前部署深度不清Pre-2026
未具名 OEM 1全球汽车 OEM未披露账户包含在“前 20 大 OEM 中 17 家”的汇总说法里低:无公开名称或项目深度Unknown
未具名 OEM 2全球汽车 OEM未披露账户包含在“前 20 大 OEM 中 17 家”的汇总说法里低:无公开名称或项目深度Unknown
未具名 OEM 3全球汽车 OEM未披露账户包含在“前 20 大 OEM 中 17 家”的汇总说法里低:无公开名称或项目深度Unknown
未具名 OEM 4全球汽车 OEM未披露账户包含在“前 20 大 OEM 中 17 家”的汇总说法里低:无公开名称或项目深度Unknown
未具名 OEM 5全球汽车 OEM未披露账户包含在“前 20 大 OEM 中 17 家”的汇总说法里低:无公开名称或项目深度Unknown

具名客户来自截至 2026 年 5 月审阅的 Applied Intuition 客户页面、合作伙伴文章、客户发布引用和政府页面。 未具名 OEM 行代表“前 20 大 OEM 中 17 家”说法中未披露的剩余部分,不应视为已识别账户。

[CU001, CU002, CU005, CU006, CU007, CU008]
FU004: OEM 渗透漏斗

Applied Intuition 的汇总 OEM 说法覆盖很广,但公开点名部分窄得多。

该漏斗突出披露收窄,而非客户弱点。从 17 降到 4 反映公开点名限制,而不是已证实的账户流失。

[CU001, CU002, CU003, CU024, CU026, CU032]
FU005: 客户时间线

公开客户证明先从卡车运输建立,再扩展到更广的 OEM 和防务可见度,随后进入当前工业自主系统垂直领域。

已审阅页面未披露确切发布日期时,日期按月份或报告生成日取整,以保持时间顺序,但不暗示精确到日。

[CU001, CU010, CU011, CU012, CU013, CU017]

6.2 国防与政府客户关系

国防是 Applied Intuition 客户基础中最能被独立佐证的一部分。公司的国防页面和面向 Army 的文章描述了它在军事自主系统和任务系统中的实际工作,但更关键的是,客户侧和政府侧页面也提到了 Applied Intuition。Army.mil 发表过 Army 与 Applied Intuition 合作的文章,Defense.gov 后来又重点引用 Secretary of the Army Daniel Driscoll 的表述,称 Applied 很快就把军用车辆做到了全自主,用于士兵测试。这个组合比普通创业公司的客户标识墙强得多,因为关系证据来自公司自有营销体系之外。 政府业务也改变了客户质量的形态。国防客户可以在很高层级验证产品能力和采购可信度,但公开证据仍让收入质量不透明。已审阅公开来源均未披露 Army 或 Air Force 业务的合同金额、上限金额、期权年限、项目范围或续约机制。因此,国防关系提高了 Applied 能销售任务关键系统的信心,但并未消除客户集中或采购周期风险。它按终端市场分散了客户基础,同时又带来了对政府预算和项目节奏的另一层依赖。 [CU010, CU011, CU012, CU013, CU014, CU015]

6.3 商用卡车与特种车辆客户

商用卡车提供了 Applied Intuition 在乘用车 OEM 之外最清晰的一部分客户证据。PACCAR 在 2020 年公开选择 Applied Intuition 用于自动驾驶卡车开发,给了公司来自大型商用车制造商客户侧的验证。Volvo Group 后来发布自己的合作公告,又增加了一个不依赖 Applied 自写营销的重卡参考客户。这两个账户重要,因为它们说明 Applied 的客户证据已经从软件验证实验室延伸到货运和工业车队运营商;在这些场景里,自动驾驶经济性绑定的是生产率、劳动力约束和正常运行时间,而不只是 ADAS 功能开发。 当前垂直页面进一步拓宽了这幅图。Applied 现在营销面向卡车、采矿、农业和建筑自主系统的专门方案,卡车和采矿页面也包含 Isuzu 与 Komatsu 的具名引述。这释放出一个信号:公司确实在向特种车队和工业自主系统扩展市场触达。与此同时,公开证据仍不均衡。这些页面证明 Applied 能在新车辆类别中赢得参考账户,但仍没有披露合同金额、车队数量、部署规模,或关系到底是试点、平台标准还是长期经常性软件项目。Nuro 作为历史案例研究出现,但现有客户文章 URL 已无法访问,削弱了证据的新鲜度和深度。 [CU017, CU018, CU019, CU020, CU021, CU022]

客户分群分析
分群估计客户数收入权重(估计)依赖风险关键客户备注
全球 OEM声称 17 家 / 具名 4 家最高Toyota;Volkswagen Group;GM;Hyundai;另有 13 个未披露 OEM 关系按客户标识数看,这是公开资料里最宽的分群;公开来源没有披露各 OEM 的 ACV 集中度或生产状态。
商用卡车2-4 家具名中高PACCAR;Volvo Group;Isuzu;Embark 衍生的卡车业务足迹客户侧引用存在,因此这是汽车以外最强的变现证据;但当前合同金额仍未公开。
国防 / 政府2 家具名中高客户包括 U.S. Army、U.S. Air Force政府证据在技术可信度和公开验证上很强,但采购节奏和合同金额未披露。
AV / 出行初创公司1 个具名公开引用中低Nuro公开证据比 OEM 和国防引用更旧、不够新鲜;部署经济性未披露。
工业自主化2 个当前具名引用新兴Komatsu;Isuzu,以及农业和建筑场景页面当前垂直页面显示业务正在向道路车辆之外扩展,但该分群收入和签约客户数未公开。

客户数和收入权重是根据具名引用及当前垂直页面综合估计,不是公司披露的分群报告。 收入权重估计表示可能的相对重要性,而非经审计贡献。

[CU004, CU017, CU018, CU020, CU021, CU027]
FU001: 客户细分组合

全球 OEM 预计贡献最高;防务和卡车运输是其次重要的收入来源。

这是估算的经济权重视图,不是公司披露的细分收入。它用于可视化可能的集中度,而非呈现经审计的收入分布。

[CU020, CU027, CU028, CU036, CU037, CU040]

6.4 客户集中度与收入依赖分析

Applied Intuition 的公开客户名单显示客户标识分散度很高,但收入仍可能集中。实践中,价值最高的关系很可能只是全球 OEM 平台、重卡项目和国防账户中的一小部分。这在企业基础设施业务里很常见:一个客户标识可能只代表一张很窄的评估许可证,另一个则可能代表横跨多个项目、年度价值大得多的部署。公开材料没有披露这种分布。因此,投资人不应假设 17/20 的 OEM 覆盖会自动转化为收入均匀分布或低集中度风险。 最可能的集中格局是杠铃。一端是少数大型汽车和卡车 OEM,它们很可能代表最大的商业软件机会;另一端是美国政府业务,战略意义和技术验证价值可能很高,但也会带来项目节奏和预算暴露。公开证据没有披露 NRR、流失、GRR、平均合同期限、采购摩擦或最大客户占比。因此,正确的尽调结论不是集中度必然很高,而是仅凭公开数据无法排除集中度风险。客户标识覆盖面和跨行业拓展是正面因素;缺失的分母是收入权重。 [CU024, CU025, CU026, CU027, CU028, CU030]

客户集中度风险表
风险因素级别证据缓释措施尽调要求
OEM 集中度(前三大 OEM)“前 20 大 OEM 中 17 家”的覆盖面没有披露 ACV 组合,因此少数 OEM 仍可能主导收入向卡车、国防和工业自主化跨行业扩展要求提供前 10 大客户 ARR / ACV 以及平台层收入集中度。
政府合同中高U.S. Army 和 U.S. Air Force 引用让客户基础更多元,但可能受采购周期和预算节奏影响政府证据来自强客户侧来源,而不只是客户标识使用要求提供合同上限、履约期、选项年和重新竞标时间表。
客户流失风险未披露公开 NRR、GRR、流失或续约数据多产品栈部署后可能提高切换成本要求按分群提供队列留存、客户标识流失和扩张历史。
合同期限未知公开来源很少区分付费生产项目、评估和工具试点具名引用说明卡车和国防领域有真实部署价值要求提供每个具名客户的合同期限、年化金额和部署阶段。
续约指标新的垂直页面显示持续的市场拓展投入,但公开续约并未单独披露当前 2026 年物理 AI 页面暗示账户开发仍在继续要求提供头部客户的续约日期、NPS / 满意度和产品扩张证据。

风险级别反映承销不确定性,并非已证实客户疲弱。公开合同经济性缺失,本身就是核心尽调问题。

[CU024, CU025, CU026, CU027, CU028, CU036]
FU002: 客户关系流程

Applied Intuition 似乎先靠仿真和工具落地账户,再扩展到更深集成、自主系统和任务系统项目。

该流程是基于 Applied Intuition 公开产品和客户证据推断出的落地-扩张模型;公开来源未披露具体商业漏斗或转化率。

[CU019, CU023, CU026, CU033, CU038]

6.5 证据质量与尽调缺口

本章最好的客户证据,来自 Applied Intuition 并非唯一叙事者的来源。PACCAR 投资者关系、Volvo Group、Army.mil 和 Defense.gov 都从外部确认 Applied 确有真实客户或项目关系。这些来源显著提高了卡车和国防账户的可信度。相比之下,乘用车 OEM 叙事仍主要由 Applied 自写页面和合作公告驱动。这仍是有用证据,但证据质量低于客户发布的案例研究、采购文件或独立报道的生产部署。前 20 大 OEM 中 17 家这一数字方向上很亮眼,但它仍是锚定较早公开材料的汇总说法,而不是 2026 年新鲜的具名账户披露。 因此,主要尽调缺口集中在合同深度,而不是 Applied 是否真的有客户。投资人仍需要当前的具名账户清单、按 ARR 或 ACV 划分的最大客户集中度、续约时间表、各主要 OEM 的生产与试点状态,以及 Army 和 Air Force 业务的详细合同条款。他们还需要一份仍可访问且新鲜的 Nuro 客户证据,或确认该关系已从公开营销中撤下。简言之,公开证据足以支持“客户真实存在”的结论,但还不足以完整支撑收入耐久性或集中度风险。 [CU014, CU022, CU024, CU025, CU026, CU030]

客户引用证据质量
客户证据类型来源日期深度指标验证状态
ToyotaApplied 撰写的合作伙伴文章Unknown具名关系;无公开生产指标仅由公司来源确认
Volkswagen GroupApplied 撰写的合作伙伴文章Unknown具名关系;无公开生产指标仅由公司来源确认
General MotorsApplied 撰写的合作伙伴文章Unknown具名关系;无公开生产指标仅由公司来源确认
HyundaiApplied 撰写的合作伙伴文章Unknown具名关系;无公开生产指标仅由公司来源确认
PACCAR客户 IR 公告及 Applied 文章2020-06-09客户侧选择公告高质量交叉佐证
Volvo Group客户公告及当前卡车业务页面2022-09-01客户侧合作公告中等质量交叉佐证
U.S. Army.mil 文章、Applied 文章及国防引用2024-2026政府侧验证和现场测试表述高质量交叉佐证
U.S. Air ForceApplied 国防引用2024-2026具名客户引用,无合同细节中等质量的公司声称证据
Nuro案例页面及已失效的直接客户 URL历史 / 未知仅历史具名引用新鲜度低的证据
Komatsu / Isuzu当前垂直页面上的具名引语2026解决方案页面上的当前具名引用中等质量的公司证据

证据质量按发布方独立性、具体程度和新鲜度分级。客户发布和政府发布的引用高于 Applied 撰写的合作伙伴文章; 即使历史证据很可能存在,链接失效或缺少在线 URL 也会降低验证质量。

[CU010, CU012, CU014, CU017, CU018, CU022]
FU003: 客户证据质量

最强客户证明来自客户和政府发布的来源;留存和生产深度数据大多仍未公开。

计数强调证据深度,而非收入贡献。零表示审阅的公开来源没有披露该指标,并不表示公司没有该能力。

[CU014, CU025, CU026, CU029, CU030, CU037]
Chapter 07

07风险

7.1 监管环境与合规风险

Applied Intuition 切入的交通和国防软件,是监管最未定型的类别之一。美国联邦 AV 政策仍更依赖指南、缺陷管辖权、事故报告和 FMVSS 渐进更新,而不是一套针对无人驾驶系统的单一上市前认证制度。这给 Applied 带来混合结果:没有联邦安全强制要求 OEM 采用它的技术栈,但也没有稳定的全国审批路径来缩短客户采购周期。CRS、NHTSA 政策页面、进行中的规则制定案卷和待审众议院立法都指向同一个状态:框架仍未完成,而非已经定稿。对投资人而言,Applied 的商业节奏部分受制于它无法控制的监管演进。 州层面和卡车层面的碎片化风险更尖锐。NCSL 立法数据库显示,自动驾驶车辆法律现在已覆盖许可、测试、保险、隐私以及在公共道路运营等议题,涉及大量州;FMCSA 的 2023 年补充 ANPRM 也确认,搭载 ADS 的卡车所适用的商业机动车规则仍在推进中。因此,每当卡车客户或 OEM 必须在联邦卡车框架落定前逐州处理运营假设时,Applied 都会面临二阶暴露。美国之外,情况不是更简单,而是更复杂。公开政策来源和法律评论仍描述美国体系与欧洲 UNECE 式型式批准逻辑之间的方法分歧;NIST 的 AI RMF 虽然仍属自愿,但随着 NIST 增加 2026 年关键基础设施剖面,它正明显成为采购参考点。再加上国防暴露,合规栈进一步扩张:EAR 出口管制、国防业务可能邻近 ITAR、FAR/DFARS 采购条款、外国收购限制和涉密项目处理,都会提高执行成本,却不保证带来收入拉动。[CR001, CR002, CR003, CR004, CR005, CR006]

监管 / 法律风险登记表
风险类别严重性发生概率监管 / 法律依据可用缓释措施剩余风险
NHTSA 无联邦强制要求监管NHTSA 政策仍偏指导性;没有联邦 AV 认证强制要求迫使 OEM 采用Applied 受益于近期合规摩擦较低,也能卖进多种 OEM 策略高:缺少强制要求时,即使规则持续变化,客户采购周期仍可能偏慢
FMCSA 规则制定待定监管FMCSA SANPRM 确认,配备 ADS 的 CMV 规则仍在审议中Applied 可在客户摸索临时合规路径时支持试点和开发项目高:卡车商业化仍可能被悬而未决的联邦要求拖慢
EU 框架分化监管中高CRS 和法律评论显示,美国与 UNECE 风格 AV 框架尚未统一多区域产品设计和客户定制合规工作可缩小差距中高:区域差异推高定制成本,并拖慢跨境落地
ITAR / 出口管制法律EAR 许可、国防业务邻近 DDTC 监管,以及外国最终用途审查,可能适用于自主系统软件和服务分类审查、筛查和许可流程可降低违规风险高:每一个新的国防或海外项目都可能重启合规分析
AV 产品责任法律中高若无联邦责任法规,AV 事故仍会进入州侵权和产品责任理论合同安排、赔偿责任分配和严格验证证据可降低但不能消除风险敞口高:重大事故中,Applied 仍可能与 OEM 和运营方一起被列名
开源冲击竞争中高CARLA 提供免费的仿真基线,且开发者采用活跃Applied 靠工作流集成、企业支持和适配国防的部署场景差异化中:只针对仿真的预算仍面临真实价格压力
CEO 关键人风险运营中高Applied 仍带有强创始人色彩,外部叙事也与 Qasar Younis 绑定工程领导层扩大和广泛招聘降低但不能消除创始人集中度中高:领导层交接可能扰动融资和战略叙事
AV 行业整合市场Argo、Embark、Motional 等案例显示,大规模商业化前行业已先经历出清Applied 的资本强度低于车队运营方,客户也更多元高:行业挫折仍会压低客户支出和市场情绪
客户集中度市场“前 20 大 OEM 中 17 家”的覆盖面没有披露收入权重或期限结构向国防和卡车等垂直方向延伸,提供了一定多元化高:少数大客户仍可能主导经济性
收入不透明 / 未披露财务公开材料披露估值和盈利能力说法,但未披露收入、ARR 或利润率私人公司纪律和投资人质量只能部分抵消高:投资人无法独立验证当前估值叙事的持久性
仿真到现实落差技术中高RAND 和 NHTSA 来源显示,验证负担极高,且仍需要现实世界监督Applied 可为客户提升工具、覆盖率和流程证据高:没有任何模拟器能完全消除边缘案例或部署迁移风险

严重性和发生概率是基于截至 2026 年 5 月审阅的公开监管、法律、行业和公司来源综合得出的定性判断。 本登记表按剩余投资风险排序,并不表示任何单项当前正在造成已披露事故。

[CR001, CR005, CR006, CR008, CR011, CR017]
监管格局表
司法辖区机构规则 / 政策状态影响级别Applied Intuition 风险敞口
美国NHTSA联邦自动驾驶车辆政策 / FMVSS 更新已生效但不完整影响客户部署节奏、报告预期和未来联邦 AV 要求
美国FMCSA配备 ADS 的 CMV 规则制定(2023 SANPRM)待定影响自动驾驶卡车客户和商业化时间表
欧盟UNECE WP.29 框架型式认证和车辆监管路径分化演变中中高抬高全球 OEM 项目的定制和认证负担
美国DoD / DDTCITAR 相关国防技术管制随项目而定可能限制境外访问、协作和国防相关自主系统软件出口
美国BISEAR 出口管理和许可生效中国际软件和技术转移需要分类、筛查,并可能需要许可证
中国MIIT / 市场准入制度本地监管审批和数据本地化敏感性不透明 / 演变中中高限制把美国或欧盟商业化假设直接复制到中国 OEM
全球SAE International自动化等级和工程标准参照点现行参考标准即便本身不是有约束力的监管机构,也会影响客户和监管沟通框架
美国NISTAI 风险管理框架自愿采用,新增 2026 年关键基础设施专项框架可能成为国防和安全敏感部署的采购预期

这张格局表概括与 Applied Intuition 公开风险边界最相关的机构。“状态”反映截至本章运行日已审阅的公开证据,并不主张任何单一机构目前直接监管 Applied 的所有产品。

[CR001, CR005, CR006, CR007, CR008, CR009]
FR001: 风险热力图

严重度与可能性合成评分显示,监管不完整、客户集中、责任归属和收入不透明是最重的残余风险。

评分是定性综合判断,不是经审计的风险指标。图中呈现章节发现的相对权重,而非数值化企业风险模型。

[CR001, CR005, CR011, CR017, CR021, CR027]
FR002: 监管合规流程

Applied 的合规负担从产品范围延伸到 AV 政策、卡车规则、出口管制和国防采购要求。

[CR005, CR008, CR009, CR010, CR040]

7.2 法律与责任风险

Applied Intuition 的法律风险,核心在于自动驾驶软件从开发工具进入运营栈后,责任会落到哪里。Greenberg Traurig 的 2026 年责任综述把要点说得很清楚:在没有专门联邦责任法的情况下,AV 事故仍会回到州法拼图、过失诉讼和产品责任理论。这很关键,因为 Applied 在自己的公开叙事中已不再只是仿真器供应商;它现在谈 Vehicle OS、SDS、国防自主系统和 physical-AI 工作流,这些都可能更接近客户部署。严重事故发生时,原告不需要像上市公司披露那样完整的证据链,就可以把每一个看似可能的被告都列入诉讼。OEM、ADS 供应商、远程运营商、软件供应商和组件合作伙伴都可能进入同一个案件,再由法院分配责任。 知识产权与执法侧同样值得重视。Waymo v. Uber 商业秘密案持续提醒市场:前员工、模型诀窍或自研开发工具发生争议时,自动驾驶软件会以很高金额进入诉讼。已审阅来源没有发现 Applied 公开具名的 AV 事故诉讼或政府同意令;这好过已披露诉讼,但缺席本身不消除风险。Husch Blackwell 对已关闭 Waymo 调查的总结显示,监管方可能花很长时间调查安全问题,最后在没有系统性发现的情况下结案。与此同时,出口管制暴露不是一次性打勾:BIS 和 Trade.gov 都描述了 EAR 下持续的许可、筛查和最终用途审查义务;国防项目还可能对数据共享、外籍人员访问和技术转让提出类似限制。一家公司若同时服务全球 OEM 和美国国防买方,法律边界可能比收入披露扩张得更快。[CR011, CR012, CR013, CR014, CR015, CR016]

7.3 技术与产品风险

最重要的产品风险,是再复杂的仿真也无法完全弥合真实世界边缘案例缺口。RAND 的安全研究仍是经典警示:要证明 AV 安全性在统计上显著优于基准,可能需要数千万到数百亿英里的里程,取决于被衡量的事件。这不是说 Applied 的仿真产品没有价值;而是说自动驾驶的举证负担内生地巨大,任何供应商都不应被按“合成环境和回放环境已经消除真实验证风险”来承销。NHTSA 自己的 ADS 研究和安全页面仍强调测试、评估、监督和事故报告,而不是暗示验证问题已经解决。对 Applied 来说,任何看似穿过仿真工作流的高知名度客户失败,都会重击信誉,即便最终法律责任落在 OEM 身上。 开源和执行风险叠加在这条技术基线之上。CARLA 为开发者提供了可信的零许可证仿真环境,带 API、场景工具和大型公共社区,因此 Applied 必须靠工作流集成、企业支持、安全流程严谨度和领域覆盖来守住价格与差异化,而不能只靠仿真本身。这是一条有意义的护城河,但并非坚不可摧。关键人物依赖是另一项主要产品执行风险。Applied 在资本市场和战略叙事中仍与 CEO Qasar Younis 高度绑定;招聘页面也显示,它仍在争夺稀缺的自动驾驶、机器人和具备国防许可背景的工程人才。国防项目还带来纯商业同行没有的约束,包括安全许可瓶颈和出口管制下的协作限制。随着公司从工具扩展到更广义的机器软件,路线图排序失误或高级领导层流失会比它还是一家更窄的仿真供应商时更重要。[CR017, CR018, CR019, CR020, CR021, CR022]

7.4 市场与商业模式风险

Applied Intuition 受益于向自动驾驶卖“铲子”,而不是直接给 robotaxi 或自动驾驶卡车车队烧钱,但行业风险并未消失。AV 行业史上不缺资金雄厚、最后仍在经济性跑通前失败、重组或被出售的项目。Embark 是最接近的警示案例,因为 Applied 自己是在这家独立卡车公司耗尽现金跑道后买下其资产。IEEE Spectrum 更广泛的行业报道和 RAND 对商业化的谨慎判断指向同一个方向:商业化耗时长于乐观投资人的预期,安全事故可以重置时间表,即便技术可信的项目也可能走向整合,而非成为独立赢家。Argo AI、Motional、TuSimple 和 Zoox 分别展示了不同失效模式——关停、收缩、治理压力或战略吸收。 Applied 还背负公开来源无法解答的商业模式不确定性。公司宣传了广泛 OEM 覆盖、国防进展、盈利能力和 150 亿美元估值,却仍未公开披露收入、ARR、毛利率、最大客户集中度或合同期限。因此,外部无法判断公司是真的不受漫长 AV 采用曲线拖累,还是只是资本更充足、能更久地对抗它。即便客户标识很多,客户集中度仍是真实风险:前 20 大 OEM 中 17 家并不说明是否有 3 个账户主导经济性。政府业务又叠加第二层集中度,因为项目节奏取决于预算、采购周期和国家安全优先级。地缘政治冲击——从出口限制到汽车供应链扰动——也可能影响 Applied 依赖的同一批客户,而这些客户最终要贡献生产级收入。相比失败的车队运营商,这门生意更有韧性;但它仍显著高于普通企业软件的风险。[CR024, CR025, CR026, CR027, CR028, CR029]

自动驾驶行业失败案例
公司失败类型融资金额关闭 / 事件日期原因Applied Intuition 启示
Argo AI关闭 / 资产收缩第三方报道获 Ford 和 VW 数十亿美元支持2022-10商业化时间表和资金需求跑在战略支持前面仅有 OEM 背书,无法保证自动驾驶公司独立且持久
Embark Trucks上市公司崩盘 / 资产出售第三方报道融资数亿美元,另有 SPAC 资金2023-08资金消耗、卡车商业化延迟和市场融资窗口疲弱Applied 可以低价买资产,但拖垮 Embark 的行业经济账仍然重要
Motional收缩 / 多次重组第三方报道获得 Hyundai 和 Aptiv 大额支持2024-2025自动驾驶出租车时间表和单位经济性仍然艰难,即便背后有强大股东即便资金充足的自动驾驶团队,也可能被迫暂停或收窄野心
TuSimple治理和战略瓦解第三方报道具备公开市场和风险资本基础2023-2024治理争议、监管审查和地缘政治复杂性削弱了投资逻辑自动驾驶卡车敞口不只考验技术,还带来治理和管辖权风险
Zoox战略收购,而非独立扩张第三方报道以十亿美元级价格出售给 Amazon2020-06独立扩张让位于更大平台公司的战略所有权技术可信的自动驾驶项目,最终价值也可能主要落在战略收购方手里
Anthony Levandowski / Uber ATG商业秘密诉讼 / 声誉受损和解,而非经营失败2018-02自动驾驶专有技术的 IP 争议带来九位数后果和战略扰动自动驾驶人才流动不能消除商业秘密和 IP 污染风险

金额刻意按公司声称或第三方报道的背景呈现,而非审计数字。表格用于说明失败模式和战略启示,不是逐一重述每个可比公司的精确资本化历史。

[CR024, CR025, CR032, CR033, CR034, CR035]
FR003: AV 行业失败时间线

行业历史显示,大规模商业化之前,关停、重组和并入战略方反复出现。

[CR024, CR032, CR033, CR034, CR035, CR036]
FR004: 风险类别分布

本章把多数残余暴露先压在监管和法律类别,其后才是技术、市场和财务后续风险。

数值代表本章强调的主要风险主题数量,不代表公司披露的暴露权重。

[CR001, CR011, CR017, CR024, CR028]

7.5 风险缓释评估

Applied 确实有实质缓释因素。它不是以单一项目车队运营商的方式进入市场;它有广泛的 OEM 相关性、军民两用的国防角度,以及一套工作流级产品叙事,这比单点仿真器更难替代。Embark 收购给了它卡车资产,又不必承接 Embark 的公开市场烧钱压力;国防业务也增加了第二条需求向量,且不与乘用车周期完全相关。公司的公开定位还显示,它理解合规导向流程的重要性:它公开谈国防用例,为受控环境构建产品,并销售给本身就必须关心安全案例、验证证据和采购纪律的客户。这些都是有意义的优势。 但剩余风险仍只被部分缓释,因为最难的问题恰恰披露最少。仅凭公开来源,投资人看不到客户集中度、合同结构、诉讼准备金假设、出口分类流程、DoD 合同工具细节,或 Applied 合规基础设施的真实深度。这意味着公司内部可能把这些风险管得很好,但外部承销人仍无法核验。实际含义是,Applied 应被视为强劲但尚未去风险化的基础设施公司。最可信的投资逻辑破裂信号包括:公开 AV 责任诉讼点名 Applied、重大监管收紧放慢客户部署却没有创造强制需求、失去大型国防或 OEM 关系,或出现证据表明 150 亿美元叙事跑在未披露经营基本面之前。在收入质量和合规深度披露更充分之前,风险缓释应评为混合,而不是完整。[CR038, CR039, CR040, CR041, CR042, CR043]

FR005: 风险缓释状态

缓释力度在工作流广度和市场地位上最强,但在公开财务透明度和合规披露深度上最弱。

KPI 是基于本章公开证据得出的定性缓释评级,不是管理层提供的控制测试。

[CR038, CR039, CR041, CR042, CR044]
Chapter 08

08估值

8.1 估值历史与市场背景

Applied Intuition 已从后期汽车工具创业公司,转向范围更宽的 physical-AI 叙事;估值轨迹也反映了这次重新定位。公开轮次数据在历史后段很清楚:2020 年 Series D 据报道融资 1.75 亿美元、估值 36 亿美元;2024 年 10 月轮次据报道融资 2.5 亿美元、估值 60 亿美元;最新 Series F 被广泛报道为融资 2.5 亿美元、估值 150 亿美元。这意味着,从 Series D 到最新轮的公开估值上调约 4.2 倍,其中从 60 亿美元到 150 亿美元的跃升尤其陡峭,在约一年内提升了 2.5 倍。 更难判断的是,市场背景能否支撑这种加速。管理层把最新轮与盈利、三位数增长、BlackRock 支持和 OpenAI 合作绑定在一起,这些都把故事从狭义仿真软件倍数推向战略基础设施框架。不过,分析师市场报告仍显示,纯 AV 仿真如今只是低个位数十亿美元市场。因此,估值要求投资人相信 Applied 正在汽车和国防领域变现更宽的自动驾驶技术栈,而不只是赢下仿真细分市场份额。[CV001, CV002, CV003, CV004, CV005, CV006]

融资轮次估值轨迹
轮次日期融资额($M)投后估值($B)收入倍数(估计)隐含收入估计倍数背景
Series A 轮2018未披露未披露N/AN/A早期轮次被创业数据库追踪,但清晰投后细节未公开。
Series B 轮2019未披露未披露N/AN/A公开证据支持融资持续推进,但没有清晰估值标记。
Series C 轮202040未披露N/AN/A后续估值跃升前的成长资本轮;投后估值未清晰公开。
Series D 轮20201753.6无法从公开数据估算未披露首个后期公开估值标记,且战略投资人支持强。
Series E 轮20242506.0~25x-30x~200-240如果收入已成规模,则符合高溢价成长软件定价。
Series F 轮202525015.0~20x-30x~500-750需要强得多但未披露的基本面,或国防加物理 AI 溢价。

Series D 之前的早期轮次估值在公开来源中不完整;后续隐含倍数来自情景化 ARR 反推,而非公司披露指标。

[CV001, CV002, CV003, CV007, CV008, CV009]
FV001: 估值轨迹

公开里程碑显示,从 2024 年末到最近一轮,估值快速上冲。

Series A 至 Series C 的时间点主要强调融资顺序,因为 Series D 之前公开的投后估值并不一致。

[CV001, CV002, CV003, CV007, CV008, CV009]

8.2 可比公司分析

Applied Intuition 没有完美的上市可比公司,因此估值纪律只能靠三角测量。Mobileye 是最干净的公开参照,因为它是有规模的自动驾驶和 ADAS 软件平台,具备公开文件和市值;如果 Applied 配得上 Mobileye 式的头部估值,投资人就应期待看到远多于当前公开证据的收入披露。Aurora 是有用的公开反例,因为它展示了公开市场会如何严厉对待 AV 故事,即便公司有监管文件、商业化进展和公开透明度。Scale AI 是最相关的私营跨界可比公司,因为它处在更广义 AI 基础设施主题之内,在公开财务细节有限的情况下仍获得了高溢价私募估值。 其余样本有信息量但并不完美。Waymo 由母公司持有且偏 robotaxi;Waabi 和 Wayve 更早期、商业证明更窄;Palantir 更宽、更成熟,但可用于观察国防 AI 溢价;Luminar 则提醒投资人,硬件相邻的自动驾驶公司可能以很深折扣交易。关键结论不是某一行能给出精确倍数,而是:除非 Applied 已经比公开证据显示的规模大得多、变现好得多,否则 150 亿美元价格处在同业集合的高端。[CV014, CV015, CV016, CV017, CV018, CV019]

可比估值表
公司阶段估值($B)收入($M 估计)倍数国防敞口备注
Applied Intuition未上市,最新轮次15未披露;估计 200-500+N/M,或在 $500M 时约 30x溢价取决于隐藏收入规模和平台逻辑。
WaymoAlphabet 子公司30-45未披露 / 不可比N/M自动驾驶出租车和母公司持有结构,让软件倍数比较并不完美。
Aurora (AUR)上市3-5未达规模 / 规模化前N/M公开市场质疑使其价值远低于顶级未上市 AI 估值标记。
Mobileye (MBLY)上市15-201,500-1,900~8x-12x最佳头部价值公开可比,但已披露收入基数大得多。
Scale AI未上市14未披露N/M最接近的未上市 AI 基础设施可比,且有战略国防重叠。
Palantir (PLTR)上市50+数十亿美元高十几倍至约 20x不是直接自动驾驶可比,但可用于国防 AI 溢价基准。
Luminar (LAZR)上市1-3硬件主导 / 未达规模低个位数销售倍数显示公开市场会如何严厉看待自动驾驶相邻硬件故事。
Waabi未上市<5 估计未披露 / 规模化前N/M低-中更早期自动驾驶软件参照,商业验证弱于 Applied。

估值为 2026 年 5 月方向性参考点,综合公开市场区间、监管文件和私募市场报道;未上市公司收入通常未披露。

[CV014, CV015, CV016, CV017, CV018, CV019]
FV002: 可比公司位置

按公开口径估值看,Applied 已接近自动驾驶软件可比区间顶部。

公开市场价值是 2026 年 5 月的方向性区间;私有公司价值是四舍五入的分析师估计,不是谈判得出的控制权价值。

[CV015, CV017, CV018, CV019, CV020, CV021]

8.3 收入模型与倍数分析

核心估值问题是分母不透明。公开来源反复披露估值、投资人质量、盈利能力和战略合作,但没有披露收入、ARR、毛利率或留存,外部投资人无法有把握地计算历史倍数。因此,正确方法是倒推出一个合理软件倍数所隐含的收入基数。在这个基础上,如果 Applied 在 2024 年按高溢价软件假设已有约 1.5 亿到 2.5 亿美元的 ARR,60 亿美元估值就算合理。相比之下,150 亿美元估值很可能需要大约 5 亿到 7.5 亿美元的 ARR,除非投资人在承销一个更高的基础设施式倍数。 商业模式细节正是在这里变得重要。OEM 工程工具最可能是核心收入驱动,但高溢价情景取决于国防项目、车辆智能授权和更宽技术栈变现,在仿真之外贡献可观增量。盈利能力支持一种判断:Applied 可能已经比失败的 AV 同行经济性更强;BlackRock 加 OpenAI 也强化了溢价叙事。即便如此,如果收入仍低于约 2 亿美元,最新轮相对公开自动驾驶和软件参照会隐含极度拉伸的倍数。[CV024, CV025, CV026, CV027, CV028, CV029]

收入模型假设
收入驱动因素权重(估计)证据质量置信度备注
OEM 工具许可证最高最可能是核心经常性收入流,因为它契合 Applied 的历史定位和客户基础。
国防合同高且上升如果合同规模可观且可重复,足以支撑溢价,但金额仍未披露。
Vehicle OS 许可证新兴低-中对平台逻辑重要,但变现规模没有公开量化。
仿真云使用量低-中按用量计费有上行空间,但公开来源未披露消耗指标。
专业服务低-中部署周边很可能存在,但不是软件级估值的主驱动。
OpenAI 合作上行叙事弹性对故事价值具备战略重要性,但公开变现证据缺位。

权重和置信度来自公开定位和融资评论估计,并非公司披露的收入分部。

[CV024, CV028, CV029, CV030, CV031, CV032]
FV004: 估值框架

投资测算逻辑从市场广度出发,再到份额获取、收入规模、倍数,最后推导隐含价值。

图为概念图,呈现投资测算逻辑,不是实测流程转化率。

[CV010, CV011, CV012, CV013, CV024, CV029]

8.4 估值情景与敏感性分析

对一家质量看起来真实、财务披露却稀疏的公司,情景分析是最好的处理方式。乐观情景下,Applied 已经走在 ARR 达到 5 亿美元或更高的路径上,反复赢得 OEM 标准化决策,把国防可信度转化为实质合同规模,并证明车辆智能或自动驾驶栈模块正在旧式仿真工具之外变现。在这种结果下,更大收入基数上的低到中十几倍倍数,可以支撑甚至超过当前 150 亿美元标记。基准情景下,公司仍很优秀,但规模低于市场期待,ARR 或许在 2.5 亿到 4 亿美元,倍数仍有溢价但不狂热。这个结果支撑一家有意义的公司,但其最新轮看起来仍有些偏高。 悲观情景更直接。如果 Applied 经常性收入仍低于大约 2 亿美元,如果 OEM 预算放缓,或仿真与自动驾驶工具更加商品化,倍数压缩可能非常严重。Embark 不是直接可比对象,但它提醒市场:商业化不达预期时,这个行业可以出现剧烈减记。按概率加权看,公开数据支持的区间更接近低双位数十亿美元,而不是一个明显打折的入场点。[CV022, CV034, CV035, CV036, CV037, CV038]

估值情景分析
情景ARR 假设($M)倍数隐含估值($B)概率核心驱动因素需跟踪信号
乐观500-75020x-30x12-1825%OEM 和国防领域广泛采用平台标准2026 年新收入披露,以及更大的国防订单
基准250-40018x-25x5-1050%强劲但比头部叙事暗示更窄的软件业务车辆智能变现和客户标准化深度
悲观100-20010x-15x1-325%OEM 放缓、仿真商品化、倍数压缩没有实质披露显示收入已进入后期规模

情景数值是分析区间,不是管理层指引;目的在于暴露支撑当前估值标记所需的收入基数。

[CV034, CV035, CV036, CV037, CV038, CV039]
FV003: 情景分析

当前 15B 美元估值更靠近公开数据区间顶部,而不是中位。

柱形使用情景中点做视觉比较,不代表精确公允价值。

[CV034, CV035, CV036, CV037, CV041, CV042]

8.5 投资逻辑与建议

Applied Intuition 比大多数自动驾驶创业公司更值得给信用。公司看起来已经盈利,拿到了顶级投资人的钱,定位已从仿真拓宽出去,也有足够的国防相关性,可以获得许多出行同行从未拿到过的基础设施溢价。这是乐观逻辑,也解释了为什么投资人愿意把 Applied 当作 physical-AI 平台,而不是狭义车辆开发工具。若管理层最终披露的收入水平能匹配最新价格,这一轮可能显得有先见之明,而不是虚高。 但仅基于公开证据,建议仍是观察,中等信心、高风险、估值偏高。问题不是 Applied 看起来弱;问题在于,支撑 150 亿美元所需的基本面仍是私有信息,而可比公司集合仍指向估值纪律。反向逻辑是:市场在充分验证收入基数之前,已经为叙事、不透明度和 AI 稀缺性付了高价。最重要的尽调请求很简单:披露 ARR、毛利率、收入结构、留存和国防合同规模,然后再重新判断 150 亿美元标记是公平,还是只是雄心勃勃。[CV040, CV041, CV042, CV043, CV044, CV045]

投资质量评分卡
维度评分(1-10)理由核心风险可比对象
市场规模8更广的自主系统软件与国防机会足够大,能支撑规模化平台结果。仅靠纯仿真太小,无法支撑 15B。Scale AI / Mobileye
竞争护城河8顶级投资人、广泛 OEM 覆盖和国防可信度,显示公司确有战略位置。可比集合不完美,切换成本深度也未完全可见。Mobileye / Waymo
收入质量6盈利说法令人鼓舞,但收入、留存和利润率仍未公开。隐藏分母可能掩盖偏高倍数。Palantir / Scale AI
团队与创始人8多次拿到资本且拥有高端合作伙伴,意味着执行可信度强。后期阶段预期现在非常高。Scale AI / Wayve
资本效率7盈利说法相较许多自动驾驶同行更有利。如果商业化拓宽,资本强度仍可能上升。Aurora / Embark
监管风险6国防敞口和企业工具降低部分自动驾驶部署风险。更广的自主系统市场仍面临采购和监管摩擦。Aurora / Waymo
退出可选性7如果收入披露跟上,可支撑 IPO 或战略出售叙事。未上市估值标记可能已计入大部分上行。Mobileye / Palantir

评分表达投资质量,而非近期回报确定性;总分没有直接映射为买入判断,主因是价格纪律。

[CV040, CV041, CV042, CV043, CV044, CV045]
FV005: 投资评分卡

业务质量得分较好,但估值支撑落后于已披露基本面。

评分是基于证据主张集综合出的分析师判断,应作为相对投资测算输入阅读,而非机械输出。

[CV040, CV041, CV042, CV045, CV047]

附录 A: Applied Intuition 融资轮次摘要

Applied Intuition 已知完成六轮融资:Series A(约 $2M,2017,a16z)、Series B(约 $40M,2018, General Catalyst)、Series C($125M,2019)、Series D($175M,估值 $3.6B,2020)、Series E($250M,估值 $6B,约 2024), 以及 Series F(估值 $15B,约 2025)。累计融资估计超过 $1.5B;Series A、B 和 F 的确切融资额未公开披露。 [CO017, CO018]

Series F 由 BlackRock 和 Kleiner Perkins 共同领投,Fidelity 与 Lux Capital 参投。同时,OpenAI 宣布与 Applied Intuition 建立战略合作,释放出 Physical AI 与 LLM 基础设施两条投资逻辑正在汇合的信号。 [CO017]

免责声明

本报告仅用于尽调目的,基于截至 2026-05-19 的公开信息生成,不构成投资建议。所有财务估计均由间接公开信号推断;Applied Intuition 尚未披露收入、年经常性收入(ARR)或利润率数据。过往融资轮次不保证未来财务表现。

证据索引

结论
编号陈述可信度来源
CO001 Applied Intuition was founded in 2017 in Sunnyvale, California, by Qasar Younis and Balasubramanian Narayanan. SO001, SO002, SO024
CO002 Applied Intuition describes its mission as 'physical AI that moves the world,' signaling a platform ambition that extends beyond traditional AV simulation software. SO001, SO007, SO013
CO003 The company's product suite spans Self-Driving System (SDS), Vehicle OS, and tools for vehicle intelligence covering simulation, testing, and data workflows. SO001, SO003
CO004 Applied Intuition monetizes development infrastructure rather than a single vehicle program, making simulation, validation, and software-stack tooling its core commercialization layer. SO003, SO007, SO024
CO005 Applied Intuition states that it employs 1,000+ engineers, including roughly 40 ex-CTOs and 30 former founders. SO001, SO004
CO006 The company lists offices or operations across Sunnyvale, Washington DC, San Diego, Florida, Michigan, London, Stuttgart, Munich, Stockholm, Gothenburg, Bangalore, Seoul, and Tokyo. SO002, SO004
CO007 Applied Intuition claims it serves 17 of the top 20 global automotive OEMs. SO001, SO005
CO008 Public customer and partner references span passenger automotive OEMs, commercial trucking, and defense agencies, showing the platform is used across multiple vehicle categories. SO005, SO006, SO022, SO023
CO009 CEO and co-founder Qasar Younis previously worked at Google and later served as a YC partner before building Applied Intuition. SO002, SO026
CO010 Balasubramanian Narayanan was the co-founder and founding CTO, giving the company deep systems-engineering DNA from inception. SO002, SO024
CO011 Peter Ludwig is the current CTO, indicating that the technical organization has broadened beyond the founding structure as the company scaled. SO002, SO024
CO012 Applied Intuition's Series D announcement added former GM CEO Rick Wagoner to its advisory board. SO009, SO024
CO013 The same Series D announcement added former Daimler CEO Dieter Zetsche to the advisory board. SO009, SO024
CO014 The 2019 Series C raised $125 million from Lux Capital, Andreessen Horowitz, and General Catalyst, bringing disclosed total funding at that point to about $175 million. SO010, SO025
CO015 The 2020 Series D raised $175 million at a $3.6 billion valuation, led by Elad Gil with Addition and Coatue participation. SO009, SO025
CO016 Applied Intuition describes Series E as a $250 million round at a $6 billion valuation led by Lux Capital, with participation from Porsche Investments, Elad Gil, BOND, Andreessen Horowitz, and General Catalyst. SO008, SO025
CO017 Applied Intuition's latest announced financing is a Series F that values the company at $15 billion and names BlackRock and Kleiner Perkins as lead investors. SO007, SO020, SO027
CO018 Based on disclosed rounds through Series F, Applied Intuition has raised roughly $1.5 billion or more in cumulative capital, although the exact total is not public. SO007, SO024, SO025, SO027
CO019 Applied Intuition announced the acquisition of Embark assets in August 2023 at a $71 million equity value. SO012, SO030
CO020 Embark's asset base included more than 1.5 million autonomous highway miles, which materially expanded Applied Intuition's trucking and autonomy dataset. SO012, SO030
CO021 PACCAR selected Applied Intuition to support autonomous trucking development, giving the company a marquee commercial-vehicle reference. SO015, SO022
CO022 Volvo Group publicly partnered with Applied Intuition, adding independent evidence that the company is used by global truck and industrial vehicle manufacturers. SO023, SO005
CO023 Applied Intuition disclosed a Volkswagen Group partnership around vehicle software and validation, broadening its European OEM footprint. SO016, SO005
CO024 Applied Intuition has publicly highlighted Toyota as a partner or customer, supporting penetration into Japanese OEM programs. SO017, SO005
CO025 GM is a named public partner or customer, indicating continued relevance with incumbent U.S. automakers. SO018, SO005
CO026 Hyundai is another named partner or customer, adding a further large global OEM reference. SO019, SO005
CO027 Applied Intuition has publicly described the U.S. Army as a customer or active defense use case. SO006, SO014
CO028 The defense page positions Applied Intuition as serving both the U.S. Army and U.S. Air Force, suggesting a defense-adjacent autonomy software franchise rather than a purely automotive tools business. SO006, SO029
CO029 Applied Intuition's OpenAI partnership implies a roadmap that blends frontier foundation models with vehicle and robotics software infrastructure. SO013, SO007
CO030 BlackRock and Kleiner Perkins participating in the latest round indicates that the cap table has expanded beyond specialist autonomy VCs toward global asset managers and top-tier franchise investors. SO007, SO026, SO027
CO031 Applied Intuition does not publicly disclose revenue, ARR, gross margin, or profitability metrics in the accessible sources used for this chapter. SO001, SO024, SO025
CO032 Accessible public sources do not provide a full current board roster, ownership percentages, or detailed governance rights for the private company. SO002, SO024, SO025
CO033 Much of Applied Intuition's financial story remains private beyond financing milestones, forcing diligence to rely on valuation markers, partner logos, and hiring scale instead of audited operating metrics. SO001, SO004, SO024, SO025
CO034 Key-person dependence remains meaningful because Qasar Younis is both the public face of the company and a central relationship holder across investors, OEMs, and defense stakeholders. SO002, SO004, SO026
CO035 The company-claimed 17-of-top-20 OEM metric signals broad design-win penetration, but it does not disclose how many relationships are production revenue versus pilot or tooling engagements. SO005, SO024
CO036 Embark's bankruptcy and subsequent asset sale to Applied Intuition illustrate ongoing consolidation and execution risk across the autonomous-vehicle sector. SO029, SO030
CO037 Independent brand validation includes CNBC Disruptor 50 recognition and later valuation coverage from CNBC, supporting recruiting and enterprise credibility even without public financial disclosure. SO021, SO027
CO038 Applied Intuition's public materials emphasize a consistent logic chain in which software tools improve development speed and safety, helping win OEM and defense customers that in turn support larger strategic financing rounds. SO003, SO005, SO007
CO039 Named references including PACCAR, Volvo Group, Toyota, Volkswagen, GM, Hyundai, and the U.S. Army provide broader customer proof than a single showcase logo set. SO014, SO016, SO017, SO018, SO019, SO022, SO023
CO040 Applied Intuition's physical-AI framing suggests a strategic expansion from AV development tools into a broader autonomy operating layer for vehicles, defense systems, and eventually robotics. SO001, SO006, SO013
CO041 Public materials support a Series B in 2018 led by General Catalyst at roughly $40 million, but exact round economics beyond the headline sizing remain lightly disclosed. SO011, SO024, SO025
CO042 Series A is only partially visible in accessible sources, with databases and investor references pointing to an approximately $2 million 2017 seed or Series A led by Andreessen Horowitz. SO024, SO025, SO026
CO043 Applied Intuition's footprint is deliberately global rather than U.S.-only, with engineering and go-to-market presence spanning North America, Europe, and Asia to support multinational OEM programs. SO002, SO004
CO044 The combination of 1,000+ engineers and 17-of-top-20 OEM penetration implies that Applied Intuition is operating at infrastructure scale for the autonomy toolchain, not as a narrow consultancy. SO001, SO004, SO005
CO045 Accessible public sources emphasize the latest $15 billion valuation more clearly than the exact cash amount raised, making valuation the cleaner anchor than dilution for the most recent round. SO020, SO027, SO028
CO046 The Embark transaction was executed through public-company sale mechanics documented in SEC materials, giving unusually strong primary evidence for an otherwise private-company M&A milestone. SO012, SO030
CM001 Third-party market reports cluster the autonomous vehicle simulation market around roughly $2.5 billion in 2024, with an eventual $7-8 billion endpoint by 2030-2035. SM001, SM003, SM004, SM005, SM007
CM002 Published AV simulation market forecasts imply a roughly 13-20% CAGR depending on the research firm's category definition and endpoint year. SM001, SM003, SM004, SM005
CM003 MarketsandMarkets projects autonomous driving software to reach roughly $7 billion by 2035 at about a 13.3% CAGR. SM002
CM004 The ADAS market is already materially larger than pure AV simulation at roughly $33 billion in 2024 with an $80 billion-plus 2030 outlook, expanding the adjacent validation budget pool. SM006, SM019
CM005 AV simulation and validation are the closest external TAM proxies for Applied Intuition's core product stack because the company's public materials emphasize simulation, testing, and vehicle-software workflows rather than operating its own autonomy fleet. SM001, SM021, SM023
CM006 Automotive OEM simulation and testing tools likely account for roughly 60% of Applied Intuition's practical SAM because incumbent automakers dominate current software-defined vehicle and ADAS validation budgets. SM001, SM003, SM021, SM023
CM007 Commercial trucking is the next most important near-term segment because highway autonomy offers narrower operating design domains and active program demand across the Aurora and Waabi ecosystem. SM021, SM025, SM026
CM008 Defense is a distinct and potentially countercyclical adjacency for Applied Intuition because public company materials emphasize government autonomy programs alongside commercial vehicle workflows. SM020, SM022, SM024
CM009 Industrial autonomy in mining, agriculture, and construction is a plausible adjacency, but public evidence of Applied Intuition's scaled penetration there is weaker than in automotive, trucking, or defense. SM018, SM023, SM027
CM010 NHTSA does not currently operate a standalone federal pre-market certification regime for autonomous vehicles; deployment still depends on existing federal safety authorities plus state-by-state permitting. SM010, SM011, SM012, SM016
CM011 NHTSA continues to collect ADS safety and incident information, which raises compliance overhead and increases the need for scenario-based simulation and validation evidence. SM011, SM013, SM014
CM012 The U.S. federal AV framework remains fragmented across guidance, reporting orders, and state deployment rules rather than one unified national approval pathway. SM010, SM012, SM016, SM017
CM013 FMCSA-facing commercial autonomous trucking rules remain incomplete, leaving nationwide driver-out trucking deployment timing uncertain. SM015, SM016, SM017
CM014 EU and UNECE pathways are still developing and add homologation complexity for global OEM customers that need multi-jurisdiction validation workflows. SM016, SM018
CM015 Defense autonomy standards are shaped more by program-specific testing, interoperability, and mission requirements than by a single civilian AV rulebook. SM016, SM020, SM022
CM016 RAND's AV safety work supports the view that road testing alone would require extraordinarily large mile counts, making simulation a necessary complement to physical validation. SM008, SM009
CM017 OEM digital transformation and software-defined vehicle programs shift more validation activity into software and simulation before physical prototypes are mature. SM001, SM021, SM023
CM018 ADAS mandates and rising safety expectations expand validation spend even if full Level 4 passenger autonomy timelines keep slipping. SM006, SM011, SM014
CM019 Defense and government autonomy spending is a real growth driver for the category, even though public sources do not cleanly isolate the tool-layer budget. SM020, SM022, SM024
CM020 Sensor and compute cost declines have lowered the economics barrier to autonomy experimentation, broadening the universe of programs that can justify simulation investment. SM018, SM019
CM021 A data flywheel exists in autonomy tooling: more scenarios, miles, and edge cases improve simulation fidelity, which in turn makes validation platforms more valuable. SM008, SM009, SM023
CM022 Open-source alternatives and internal engineering stacks create pricing pressure at the low end of the autonomy-tooling market. SM018, SM023
CM023 The AV development-tools market remains fragmented because customers buy combinations of simulation, data, mapping, validation, and vehicle-software tools rather than one monolithic stack. SM001, SM003, SM023, SM025, SM026, SM027
CM024 Fragmentation benefits infrastructure vendors like Applied Intuition, but it also keeps procurement modular, comparison-heavy, and sensitive to pricing pressure. SM018, SM021, SM023
CM025 Commercialization timelines in autonomy are often measured in five to ten years from program start to scaled deployment, slowing revenue realization for tooling vendors. SM008, SM018, SM019
CM026 Safety incidents and public setbacks have shifted buyer attention away from generalized robotaxi exuberance toward constrained domains such as ADAS, trucking, and defense. SM006, SM018, SM019
CM027 Sector failures and consolidation show that autonomy-tool vendors sell into an ecosystem with meaningful customer mortality and periodic capital droughts. SM018, SM019
CM028 Regulatory fragmentation increases the relevance of simulation and compliance tooling, but it also lengthens both customer deployment timelines and sales cycles. SM010, SM012, SM016
CM029 No cited public market report in the accessible record discloses Applied Intuition's exact market share in AV simulation or autonomy software tools. SM001, SM002, SM003, SM005, SM006, SM007
CM030 Applied Intuition is publicly positioning defense as a major growth vector rather than a side experiment. SM021, SM022, SM024
CM031 DoD and federal autonomy buyers create a different demand profile from commercial AV customers because procurement is program-based, security-sensitive, and validation-heavy. SM020, SM022, SM024
CM032 Public federal spending data imply multi-billion-dollar annual autonomy-relevant spending, but exact contract values attributable to Applied Intuition's tool layer are not directly disclosed. SM016, SM020, SM022
CM033 Defense can smooth cyclicality from commercial AV downturns, but contract timing and classification limit transparency. SM020, SM022
CM034 A broad TAM lens combining AV simulation, autonomy software, ADAS-adjacent tooling, and defense autonomy budgets yields roughly a $40-45 billion opportunity set. SM001, SM002, SM006, SM020
CM035 A more practical SAM focused on OEM, trucking, and defense external-tooling spend is closer to roughly $10-12 billion. SM001, SM020, SM023
CM036 A rough SOM of about $3 billion is plausible for Applied Intuition's best-fit external-tooling opportunity, but it remains low-confidence without pipeline and win-rate data. SM020, SM021, SM023
CM037 The broad TAM is larger than today's realized spend because many autonomy programs remain in research, pilot, or limited deployment rather than fleet-scale production. SM001, SM018, SM019
CM038 Applied Intuition's position is strongest where validation complexity, safety evidence, and cross-domain integration matter more than owning the end vehicle program. SM021, SM022, SM023
CM039 Automotive OEM tooling remains the anchor segment because incumbent automakers must validate software-defined vehicles and increasingly advanced driver assistance even if robotaxi rollout slows. SM006, SM021, SM023
CM040 Highway trucking remains one of the clearest near-term commercialization paths for Level 4 autonomy, making it strategically important for autonomy-tool vendors. SM019, SM025, SM026
CM041 Defense is strategically attractive because mission value and logistics resilience can justify autonomy budgets even when consumer AV adoption is slow. SM020, SM022, SM024
CM042 Industrial autonomy is attractive as an adjacency, but current Applied-specific traction there is less visible than in automotive or defense. SM018, SM023, SM027
CM043 Third-party AV simulation market estimates vary materially by scope, geography, and forecast horizon, so they should be used as bounded ranges rather than a single-point truth. SM001, SM003, SM004, SM005, SM007
CM044 Accessible public sources do not disclose Applied Intuition's current revenue or ARR, preventing a direct public conversion from TAM/SAM to realized company scale. SM021, SM022, SM023
CM045 Regulatory progress is more likely to add compliance, reporting, and validation work than to eliminate the need for autonomy-development tools. SM010, SM014, SM016
CM046 Applied Intuition's market opportunity is broad enough to matter, but execution quality, procurement timing, and customer program survival still matter more than the headline TAM number alone. SM018, SM021, SM023
CP001 Applied Intuition's competitive landscape spans direct simulation vendors, adjacent data/open-source substitutes, and end-to-end AV platforms rather than one single peer set. SP001, SP004, SP008, SP009, SP015, SP019
CP002 dSPACE, ANSYS, IPG Automotive, and Cognata are the clearest direct simulation and testing peers because their public positioning centers on AV or vehicle-development tooling rather than operating one autonomy fleet. SP001, SP003, SP004, SP005
CP003 dSPACE is an established automotive test and simulation incumbent whose ASM and VEOS products map closely to HIL/SIL and validation workflows. SP001, SP002
CP004 ANSYS markets AVxcelerate as an autonomous-vehicle simulation suite, making it the closest large-scale engineering-software incumbent to Applied's simulation core. SP004, SP020
CP005 IPG Automotive's CarMaker competes in vehicle-dynamics, ADAS, and scenario-validation workflows that often sit inside OEM testing pipelines. SP005, SP025
CP006 Cognata positions OneSim and AVBox around synthetic scenario generation and simulation, making it a focused cloud-simulation challenger. SP003, SP018
CP007 Metamoto is a smaller, focused AV-simulation specialist and therefore a narrower threat than the main incumbents, but it still competes for specialized simulation workflows. SP006, SP018
CP008 VectorCAST competes more on embedded software verification and test automation than on full closed-loop AV simulation, so its overlap with Applied is real but narrower. SP007, SP020
CP009 Applied Intuition differentiates itself from point tools by marketing a broader physical-AI and vehicle-software stack spanning simulation, data, validation, SDS, and Vehicle OS. SP019, SP020, SP021
CP010 Direct simulation vendors compete hardest with Applied on validation budgets, procurement inertia, and safety-process credibility rather than on ownership of the entire vehicle stack. SP001, SP004, SP005, SP019, SP020
CP011 CARLA exerts pricing pressure because its GitHub repository and documentation provide a zero-license open-source AV simulator for research and prototyping. SP009, SP010
CP012 Scale AI competes adjacently on labeled data, evaluation, and data operations rather than on closed-loop simulation alone. SP008, SP020
CP013 Scale AI widens Applied's competitive surface because autonomy buyers frequently budget simulation, data curation, and model evaluation together. SP008, SP020, SP021
CP014 CARLA is strongest as a benchmark and prototyping environment rather than as a turnkey enterprise replacement for OEM-grade workflows. SP009, SP010, SP005
CP015 dSPACE, ANSYS, and IPG appear stronger than CARLA on enterprise support, deterministic workflow integration, and safety-process fit for large OEM programs. SP001, SP004, SP005, SP009, SP010
CP016 Applied Intuition's claim to serve 17 of the top 20 global automotive OEMs represents a distribution advantage that smaller challengers are unlikely to match quickly. SP019, SP020
CP017 Applied's defense positioning broadens the competitive map beyond civilian AV simulation, and that posture is less visible across most commercial-only competitors reviewed here. SP022, SP019, SP001, SP003
CP018 Waymo, Mobileye, Wayve, Aurora, and Waabi matter competitively even when they are not pure tool vendors because they shape OEM expectations about what autonomy platforms should look like. SP011, SP012, SP013, SP015, SP017
CP019 End-to-end AV companies compete for architecture mindshare by arguing that buyers should choose a full platform relationship rather than stitch together modular tools. SP012, SP015, SP016, SP017, SP025, SP026
CP020 Waymo is more relevant to Applied as a technical benchmark, talent magnet, and credibility reference point than as a general-purpose external simulation-software seller. SP011, SP025, SP026
CP021 Aurora and Waabi are focused enough on trucking autonomy that they are less direct point-tool substitutes than they are strategic competitors for program budget and partner attention. SP015, SP016, SP017
CP022 Wayve competes primarily through embodied-AI narrative and OEM mindshare rather than through sale of a classic simulation toolkit. SP012, SP025, SP026
CP023 Mobileye competes through incumbent OEM relationships and platform trust even though its posture is closer to licensable vehicle intelligence than to a standalone simulator. SP013, SP014, SP025
CP024 Applied's moat is strongest when the customer wants one vendor spanning simulation, data, validation, and vehicle-software integration rather than a best-of-breed point tool. SP019, SP020, SP021
CP025 The competitive set spans a very wide capital-scale range, from zero-price open source to billion-dollar platform companies, which means Applied faces both pricing pressure and scale pressure. SP008, SP009, SP011, SP012, SP015, SP017, SP023
CP026 A synthesized funding lens suggests Applied does not have a capital monopoly in the category because Scale AI, Wayve, Aurora, and Waabi are also well-capitalized, while dSPACE and ANSYS bring incumbent scale. SP001, SP004, SP008, SP012, SP015, SP017, SP023
CP027 dSPACE is Applied's strongest direct incumbency risk because entrenched automotive test workflows can be hard to displace even when a newer platform is broader. SP001, SP002, SP005
CP028 ANSYS is a credible medium threat because it combines simulation breadth with large enterprise-software budgets, though it appears less automotive-native than Applied. SP004, SP020
CP029 Cognata is a lower-scale but real niche threat in synthetic data and cloud simulation rather than a broad platform equal to Applied. SP003, SP018
CP030 CARLA is a low-medium direct threat but a persistent pricing anchor because its effective license cost is zero. SP009, SP010
CP031 Scale AI is an emerging medium threat because ownership of data labeling, evaluation, and feedback loops can let data vendors move upstream into validation workflows. SP008, SP020, SP021
CP032 Displacement risk rises if OEMs consolidate around internal or end-to-end stacks from Mobileye, Waymo-like benchmarks, or startup platforms instead of modular third-party tools. SP011, SP013, SP015, SP017, SP025
CP033 Applied's defense footprint is one of the clearest differentiators versus open-source and commercial-only rivals in this chapter's competitor set. SP022, SP009, SP010
CP034 Applied's speed-to-market positioning implies it sells workflow compression and deployment velocity, not just software modules. SP021, SP023
CP035 The broader the stack Applied sells, the broader its competitive surface becomes across simulation, data, operating-system infrastructure, and end-to-end autonomy platforms. SP019, SP020, SP021, SP008, SP013
CP036 Exact competitor revenues, ARR, and realized ACV are not publicly disclosed across most of the point-tool set reviewed here, limiting precision on price-to-value comparisons. SP001, SP003, SP004, SP005, SP006, SP007
CP037 Public sources reviewed for this chapter do not establish that Applied holds exclusive OEM contracts or durable sole-source positions, so multi-homing risk should be assumed until management proves otherwise. SP019, SP020, SP025, SP026
CP038 Because buyers can combine multiple tools or internal components, Applied's moat ultimately depends on renewal depth and workflow integration rather than simple first-purchase wins. SP020, SP024, SP025, SP026
CI001 Applied Intuition's 2017 Series A was small and is only publicly recoverable in approximate terms at about $2 million, with Andreessen Horowitz named as the early institutional backer. SI013, SI022, SI023
CI002 Applied Intuition's 2018 Series B was approximately $40 million and led by General Catalyst, but exact public round economics remain limited. SI014, SI022, SI023
CI003 Applied Intuition announced a $125 million Series C in 2019 and said cumulative funding had reached roughly $175 million. SI013, SI022, SI023
CI004 Applied Intuition announced a $175 million Series D in 2020 at a $3.6 billion valuation. SI012, SI022, SI023
CI005 Applied Intuition announced a $250 million Series E in late 2024 at a $6 billion valuation. SI011, SI021, SI009
CI006 Applied Intuition's 2025 Series F valued the company at $15 billion with BlackRock and Kleiner Perkins leading and Fidelity and Lux Capital participating. SI010, SI007, SI027
CI007 Public reporting around the Series F clusters around an undisclosed round amount estimated near $500 million rather than a company-confirmed cash figure. SI007, SI008, SI026
CI008 Because Series A, Series B, and Series F proceeds are not fully disclosed, Applied Intuition's cumulative funding is best treated as an estimate of roughly $1.5 billion or more. SI010, SI022, SI023
CI009 SEC EDGAR search results show Applied Intuition has filed Form D notices under Applied Intuition naming variants, corroborating use of private-placement filings. SI001, SI002
CI010 The two SEC search variants reduce entity-name ambiguity because both 'Applied Intuition' and 'Applied Intuition Inc' return relevant Form D results. SI001, SI002
CI011 Lux Capital appears across Series C, Series E, and later participation, making it one of Applied Intuition's most durable institutional backers. SI013, SI011, SI010
CI012 BlackRock and Kleiner Perkins entering the latest round broadened Applied Intuition's investor base from classic venture firms toward crossover and institutional capital. SI010, SI007, SI030
CI013 Porsche Investments' participation in the 2024 round added a strategic automotive-capital signal beyond purely financial investors. SI011, SI021, SI024
CI014 OpenAI's partnership appeared alongside the latest financing narrative, strengthening the interpretation of the round as a physical-AI infrastructure bet rather than an automotive-only financing. SI010, SI027, SI029
CI015 Applied Intuition said in its Series E post that the company is profitable. SI011, SI021
CI016 Applied Intuition also said in its Series E post that it was growing at a sustainable triple-digit percentage year over year. SI011, SI021
CI017 Applied Intuition does not publicly disclose revenue in the reviewed official materials or market-data profiles. SI018, SI022, SI023
CI018 Applied Intuition does not publicly disclose ARR in the reviewed official materials or market-data profiles. SI018, SI022, SI023
CI019 Applied Intuition does not publicly disclose gross margin in the reviewed official materials or market-data profiles. SI018, SI022, SI023
CI020 Applied Intuition says it has more than 1,000 engineers. SI019, SI020
CI021 Applied Intuition's public materials claim it serves 17 of the top 20 global automotive OEMs, which is a useful scale proxy but not a financial metric. SI018, SI010
CI022 The Embark asset acquisition was structured as an all-stock transaction valued at about $71 million. SI015, SI017, SI004
CI023 Applied Intuition announced completion of the Embark acquisition on August 2, 2023. SI017, SI004
CI024 Applied framed the Embark deal as a way to acquire autonomous-trucking software assets, team, and data rather than to buy a stand-alone public-company shell. SI016, SI017, SI004
CI025 Embark proxy materials provide primary evidence that the target came from a financially distressed AV-trucking context rather than a conventional growth acquisition. SI003, SI004, SI005
CI026 Embark's liquidation outcome is an adverse sector signal because even a well-funded AV-trucking company with public-market access failed to sustain commercialization. SI005, SI003, SI028
CI027 Applied Intuition's public valuation stepped from $3.6 billion in 2020 to $6 billion in 2024 and then to $15 billion in 2025 without a matching expansion of public operating KPI disclosure. SI012, SI011, SI010, SI027
CI028 The jump from $6 billion to $15 billion in roughly one year implies a 2.5x step-up before public revenue, ARR, or margin metrics were disclosed. SI011, SI010, SI027
CI029 The latest financing is easier to corroborate on valuation than on cash proceeds because public reports are consistent on the $15 billion price signal but not on exact round size. SI007, SI008, SI027, SI026
CI030 Applied Intuition has not publicly disclosed current board composition, founder dilution, or secondary-sale terms, limiting cap-table reconstruction. SI019, SI022, SI023
CI031 No public debt facility or project-finance obligation was identified in the reviewed sources, so the main financing dependency appears to remain future equity or internal cash generation. SI001, SI002, SI023
CI032 PACCAR's announced relationship is partner proof that Applied Intuition supports commercial-vehicle development workflows with monetizable enterprise use cases. SI006, SI015
CI033 Volvo Group's partnership page provides second independent partner proof outside Applied's own materials that the company has commercially relevant trucking and industrial relationships. SI031, SI016
CI034 Applied Intuition's named investor mix now spans venture, crossover, and strategic pools of capital rather than a single investor archetype. SI010, SI007, SI022
CI035 CB Insights and PitchBook help summarize Applied Intuition's valuation and funding history but do not provide audited revenue-quality metrics. SI022, SI023
CI036 The absence of public revenue, ARR, margin, cash balance, and concentration data means Applied Intuition cannot be fully underwritten from public evidence despite the profitability claim. SI018, SI022, SI023, SI011
CI037 The Series F investor slate and OpenAI partnership together suggest capital is being raised to extend the platform into broader physical-AI infrastructure rather than only automotive simulation. SI010, SI027, SI029
CI038 Peer capital-efficiency comparison is inherently rough because Waymo, Waabi, Scale AI, and Applied Intuition disclose different combinations of funding, valuation, and operating metrics. SI034, SI033, SI035, SI023
CI039 Applied Intuition sits between software-style private-company disclosure and autonomy-platform capital intensity, so both software and AV peers are needed for benchmarking. SI010, SI011, SI032, SI033, SI035
CI040 The facts in this chapter most sensitive to 2026 freshness are valuation, latest investors, profitability status, and whether any new financing or Form D filing has appeared. SI001, SI002, SI010, SI027
CI041 The key investor diligence asks are ownership percentage, board or observer rights, pro-rata rights, secondary-sale history, and exit-horizon expectations for the newest capital providers. SI010, SI023, SI030
CI042 Customer concentration and NRR are not publicly disclosed, so partner logos and OEM penetration claims cannot substitute for contract-quality metrics. SI018, SI022, SI023
CE001 Applied Intuition publicly groups Self-Driving System, Vehicle OS, and Tools for Vehicle Intelligence into one Physical AI stack for moving machines. SE001, SE002, SE005, SE015
CE002 Applied says SDS is designed to operate across land, air, and sea and across roads, mines, farms, and complex operational environments. SE003, SE011
CE003 Applied describes SDS as combining domain-specific sensing, compute, and control architectures rather than one identical stack for every machine. SE003, SE006
CE004 Applied says real-world deployments across vehicles, trucks, mining machines, and other hardware collect petabytes of sensor data that feed a compounding data flywheel. SE003, SE005
CE005 The automotive SDS offering is positioned as an end-to-end ADAS and autonomy stack with a path from L2+ to L4. SE006, SE015
CE006 The trucking surface describes a customized end-to-end autonomy stack that can work with any hardware and is built jointly with customer teams. SE007, SE015
CE007 Applied has live off-road product pages for agriculture, construction, and mining, showing the SDS thesis has been extended beyond passenger vehicles. SE008, SE009, SE010
CE008 Vehicle OS is presented as a single platform spanning domains and unifying perception, planning, controls, and other critical machine systems. SE004, SE001
CE009 Applied says Vehicle OS can reduce cross-domain integration effort by up to five times. SE004, SE001
CE010 Vehicle OS describes code-first workflows in which engineers model machine behavior in Python, manage changes through pull requests, and rely on built-in observability. SE004
CE011 Applied says Vehicle OS integrates on-board software, off-board services, cloud tooling, and hardware platforms in one development environment. SE004, SE002
CE012 Applied says virtualized testing can shift validation earlier and compress testing timelines from months to days. SE004, SE002
CE013 Tools for Vehicle Intelligence is described as supporting petabyte-scale ingestion, curation, and processing across fleets and long-running programs. SE005, SE002
CE014 Applied says the platform includes an SDK and modular primitives so teams can build workflows across cloud, on-prem, and air-gapped environments while bringing their own models, simulators, and metrics. SE005, SE004
CE015 Applied says the Physical AI platform turns real-world sensor data into curated, labeled segments for downstream training, simulation, and evaluation in a closed loop. SE005, SE014
CE016 The Physical AI page explicitly says the system is designed to be orchestrated by agents and exposes tokenized UI designs plus MCP-ready interfaces. SE005, SE015
CE017 Applied's research surface highlights world-action foundation models, vision-language-action work, 4D reconstruction, and reinforcement-learning post-training for physical AI. SE014, SE015
CE018 Applied says its research organization is supported by large-scale neural simulation, synthetic data, and ML infrastructure scaling to thousands or more GPUs. SE014, SE015
CE019 Applied's defense page says the company builds autonomy software, simulation infrastructure, and mission systems for contested environments across domains. SE011, SE003
CE020 Applied's defense page markets collaborative autonomy over mesh networks and multi-machine coordination as a defense differentiator. SE011
CE021 In the Army ISV story, Applied says it turned an Infantry Squad Vehicle fully autonomous in 10 days and paired it with a Humvee mobile command post. SE012, SE011
CE022 Applied says the Army field exercise used off-road autonomy software and Vehicle OS together for onboard and remote control, route definition, progress monitoring, and hazard alerts. SE012, SE004
CE023 Applied's public narrative presents defense and commercial products as a dual-use loop in which mission-critical rigor feeds commercial products and vice versa. SE011, SE015
CE024 Applied's current case-study index spans 2023 through 2026 and includes automotive, trucking, and defense-adjacent customer proof rather than a single-domain customer story. SE013, SE007, SE011
CE025 Compared with CARLA, Applied's disclosed differentiation is breadth because it sells OS, autonomy, data workflows, and domain-specific deployment services rather than only an open-source simulator. SE002, SE003, SE004, SE016, SE017
CE026 CARLA is a credible open-source research baseline with a Python API, scenario tooling, ROS bridge, and OpenDrive support, but it still requires substantial local build and infrastructure work. SE016, SE017
CE027 Ansys AVxcelerate and CarMaker both market closed-loop simulation, SiL or HiL coverage, and safety or approval workflows, showing that Applied competes against mature validation suites rather than only startups. SE018, SE019
CE028 Ansys explicitly markets ASAM alignment, homologation support, sensor-accurate simulation, and safety justification, while Applied's public pages claim compliance outcomes without naming the standards stack. SE018, SE002, SE004
CE029 CarMaker explicitly markets MIL, SIL, HIL, and VIL coverage plus supported standards and approval workflows, which is a more transparent public standards posture than Applied currently provides. SE019, SE002, SE004
CE030 Mobileye publicly names RSS, REM mapping, and ISO 9001 on its about page, showing that some platform competitors disclose specific technical primitives and certifications more explicitly than Applied does publicly. SE023
CE031 Waymo Open Dataset and the 1001 Hours paper illustrate the scale expectations around autonomy data and simulation research that Applied's petabyte and research-infrastructure narrative is trying to satisfy. SE021, SE022, SE005, SE014
CE032 Public Applied pages support code-first, SDK-based, and cloud or air-gapped workflow claims, but they do not publicly document exact REST endpoints, SDK package names, or benchmark throughput. SE004, SE005
CE033 Applied's product pages claim compliance with regulatory standards and safety validation, yet no reviewed source named ISO 26262, ISO 21448, MISRA, AUTOSAR, or ASAM compliance for Applied specifically. SE003, SE004, SE006, SE011
CE034 The observable AWS Marketplace path plus Applied's code-first and SDK messaging suggest cloud-adjacent distribution intent, but marketplace details were not publicly inspectable in this run. SE025, SE004, SE005
CE035 Applied's public product narrative repeatedly promises faster deployment through modular OS, virtualized testing, and reusable autonomy software, framing value as weeks or days instead of longer integration cycles. SE004, SE011, SE015
CE036 Applied's disclosed stack is modular enough to cover passenger vehicles, trucks, defense vehicles, mines, farms, and construction fleets while retaining a common software and control substrate. SE003, SE007, SE008, SE009, SE010, SE011
CE037 The clearest technical moat in public materials is not a single algorithm but the combination of petabyte-scale data operations, closed-loop simulation, code-first OS tooling, and domain-specific deployment services. SE003, SE004, SE005, SE014, SE015
CE038 The clearest public product gaps are missing named standards or certifications, undisclosed simulation throughput, and unclear third-party HIL or AUTOSAR integration details. SE004, SE018, SE019, SE020
CE039 SambaNova's 2025-2026 blog archive shows that agentic inference and MCP remain current infrastructure themes, making Applied's 2025-2026 MCP-ready language directionally timely rather than stale. SE024, SE005
CE040 Applied's product pages present SDS and Vehicle OS as deployable software foundations for customer machines rather than as a branded end-to-end vehicle product of its own. SE001, SE002, SE003, SE006
CE041 The current public surface describes simulation, evaluation, ingestion, quality control, and data collection inside Tools for Vehicle Intelligence rather than foregrounding older Simian or Spectral brand names. SE002, SE005
CE042 Those simulation and data-management functions still clearly exist on the current product surface, so legacy Simian and Spectral capabilities appear to have been absorbed into the broader Tools for Vehicle Intelligence umbrella. SE002, SE005
CE043 Scale AI positions itself as a data platform for AI, so it is a useful benchmark for data operations but not for full vehicle simulation or OS control. SE026, SE005
CE044 In the 2025 Series F post, Applied said it had set out eight years earlier to accelerate the world's adoption of safe, intelligent machines, anchoring the current stack in a 2017 start. SE015
CE045 The current case-study index shows public customer-proof entries in 2023, 2025, and 2026, including May Mobility, Toyota, Isuzu, Scientific Systems, and AISIN. SE013
CE046 The Series F post says recent product launches include modular Vehicle OS and advanced autonomy stacks that are gaining traction with OEMs and fleet operators. SE015
CE047 Applied's 2026 research page publicizes ICLR 2026 and CVPR 2026 work on world models, VLA post-training, and closed-loop simulation. SE014
CU001 Applied Intuition publicly claims it serves 17 of the top 20 global automotive OEMs. SU001, SU003, SU023
CU002 Publicly named passenger-vehicle OEM references include Toyota, Volkswagen Group, GM, and Hyundai. SU001, SU005, SU006, SU007, SU008
CU003 Public sources do not identify the remaining OEM names inside the 17-of-20 claim. SU001, SU003, SU025, SU026
CU004 Applied's public customer surfaces span passenger automotive, defense, trucking, mining, agriculture, and construction, indicating a multi-vertical customer base rather than a single-segment footprint. SU001, SU002, SU015, SU017, SU020, SU021, SU022
CU005 Toyota is a named public customer or partner reference, supporting Applied's penetration into Japanese OEM programs. SU005, SU001
CU006 Volkswagen Group is a named public customer or partner reference, supporting Applied's European OEM footprint. SU006, SU001
CU007 GM is a named public customer or partner reference, indicating continued relevance with a major U.S. automaker. SU007, SU001
CU008 Hyundai is another named public customer or partner reference, adding a further large global OEM account. SU008, SU001
CU009 The customers and case-studies pages together support broad customer proof beyond a simple logo wall, although they still do not provide account-level economics. SU001, SU002
CU010 The U.S. Army is a named public customer or active program user for Applied Intuition. SU012, SU013, SU014
CU011 Army-side sources describe Applied Intuition in predictive-logistics and autonomous-vehicle contexts, showing defense work that goes beyond a static logo reference. SU013, SU018
CU012 Applied's public defense materials indicate that the company works with the U.S. Air Force as well as the Army. SU015, SU004
CU013 Applied's defense-tech materials highlight a rapid path from autonomous conversion to soldier testing, implying unusually fast customer deployment cycles in defense programs. SU016, SU014
CU014 Defense customer proof is materially stronger than most passenger-OEM proof because it includes customer-side and government-side sources rather than only Applied-authored partnership pages. SU013, SU014, SU016
CU015 Reviewed public sources do not disclose contract value, duration, option years, or renewal terms for Applied's Army or Air Force work. SU012, SU013, SU014, SU015
CU016 Government relationships diversify Applied's customer base by end market but also introduce procurement-cycle and budget-timing risk. SU013, SU014, SU025
CU017 PACCAR publicly selected Applied Intuition for autonomous truck development in 2020. SU009, SU010, SU036
CU018 Volvo Group publicly partnered with Applied Intuition, adding customer-side proof for a major commercial-vehicle relationship. SU011, SU017
CU019 Applied's current trucking page markets solutions from L2 driver assistance to full L4 autonomy and includes a named Isuzu quote, indicating active commercial-fleet expansion. SU017, SU009
CU020 Applied's mining page includes a named Komatsu quote, indicating current industrial-autonomy customer proof beyond road vehicles. SU022, SU019
CU021 Agricultural and construction pages show that Applied is actively marketing into adjacent off-highway fleets, but named public customer depth in those categories is still thin. SU020, SU021
CU022 Nuro appears as a historical customer or case-study reference in Applied surfaces, but the direct nuro-customer URL is no longer live, leaving freshness and deployment depth unclear. SU002, SU032
CU023 Applied's trucking footprint deepened partly through the Embark asset acquisition rather than only through organic customer wins. SU003, SU017
CU024 Because the remaining OEM names are undisclosed, public evidence cannot map exact account concentration inside the 17-of-20 claim. SU001, SU003, SU025, SU026
CU025 Reviewed public sources do not disclose NRR, GRR, churn, or average contract length for Applied Intuition's customers. SU025, SU026, SU023
CU026 For most named OEM relationships, public materials do not distinguish clearly between pilot work, validation tooling, and production deployment. SU001, SU002, SU005, SU006, SU007, SU008
CU027 Customer-count breadth does not equal revenue diversification because OEM, trucking, defense, and industrial programs likely carry very different annual contract values. SU001, SU025, SU026
CU028 The most revenue-relevant named customers are likely a small set of major OEMs plus government programs, implying plausible concentration risk despite broad logo coverage. SU001, SU012, SU015, SU025, SU026
CU029 PACCAR and Volvo provide independent non-company proof that Applied has monetizable heavy-vehicle customers beyond passenger-car OEMs. SU010, SU011, SU017
CU030 Customer proof quality is strongest when the source is customer-issued or government-issued rather than an Applied marketing page. SU010, SU011, SU013, SU014
CU031 The 17-of-top-20 OEM statistic traces back to older public materials, while fresher 2026 customer evidence is concentrated in defense, trucking, and industrial-autonomy pages. SU003, SU016, SU017, SU022
CU032 The 17-of-top-20 OEM figure may still be directionally useful in 2026, but it is not a fresh named-account disclosure and therefore should not be treated as current account-quality proof. SU003, SU001, SU024, SU025, SU035
CU033 The OpenAI partnership is strategically relevant to customer expansion because it positions Applied as broader physical-AI infrastructure rather than only a simulation vendor. SU023, SU031, SU016, SU034
CU034 Independent defense trade coverage shows Applied's government customer story is part of the broader market narrative rather than a one-off company post. SU027, SU028, SU016
CU035 Public materials support customer breadth across automotive, defense, trucking, and industrial autonomy, but they do not provide segment-by-segment customer counts or revenue denominators. SU001, SU015, SU017, SU020, SU021, SU022
CU036 PitchBook, CB Insights, and constrained independent coverage all point to substantial customer-relationship opacity despite Applied's visible scale and valuation. SU024, SU025, SU026, SU035
CU037 Without public ACV, renewal, and expansion metrics, customer quality cannot be fully underwritten from logo breadth alone. SU001, SU025, SU026
CU038 Applied's public customer motion appears to land with tooling and validation, then expand into deeper autonomy, fleet, and mission-system programs. SU016, SU017, SU022, SU031
CU039 Named Isuzu and Komatsu references on current vertical pages show customer proof beyond the older public passenger-OEM set, although those pages do not disclose contract value or fleet scale. SU017, SU022
CU040 Public evidence is strong enough to conclude Applied has real customers, but not strong enough to determine top-customer share, renewal health, or full revenue durability. SU001, SU024, SU025, SU026
CR001 NHTSA's federal AV policy remains guidance-heavy rather than a binding federal certification mandate for autonomous-vehicle deployment. SR001, SR004
CR002 CRS and NCSL both indicate that autonomous-vehicle governance is still split across federal and state layers, leaving material state-by-state fragmentation in testing, licensing, insurance, and operations. SR004, SR008
CR003 The continued relevance of H.R. 3935 and related federal AV legislation underscores that Congress has not yet settled a comprehensive national framework for driverless systems. SR005, SR017
CR004 NHTSA and related legal commentary show that U.S. AV policy is still evolving through incremental FMVSS updates and rulemakings rather than through one complete autonomous-vehicle approval regime. SR006, SR015, SR016
CR005 FMCSA's 2023 SANPRM confirms that the agency is still considering how to amend motor-carrier rules for ADS-equipped commercial vehicles, so trucking compliance costs and timing remain unsettled. SR007, SR004
CR006 For a vendor like Applied Intuition, the U.S. framework remains materially different from Europe's more type-approval-oriented approach, increasing cross-region compliance complexity for global OEM programs. SR004, SR015, SR017
CR007 NIST's AI Risk Management Framework is still voluntary, but the 2026 critical-infrastructure profile shows it is becoming a more concrete reference point for safety-sensitive AI procurement. SR009, SR012
CR008 BIS and Trade.gov make clear that U.S. export controls on software and technology are recurring licensing and screening obligations, not a one-time classification exercise. SR010, SR011
CR009 FAR and DFARS add procurement, foreign-acquisition, privacy, and data-rights burdens to defense software programs, raising the compliance load on Applied's defense business. SR012, SR013
CR010 China-related market access for advanced autonomy software is likely harder than Western-market deployment because export-control scrutiny and local regulatory expectations can both constrain delivery models. SR010, SR011, SR004
CR011 Absent a dedicated federal AV liability statute, autonomous-vehicle crashes still default to patchwork state law and conventional tort and product-liability theories. SR014, SR004
CR012 In Level 4-style deployments, OEMs, ADS providers, remote operators, and component suppliers can all be named in litigation, so a software stack vendor like Applied could be pulled into multi-defendant suits. SR014, SR019
CR013 Applied's public expansion from tooling into Vehicle OS, SDS, and defense autonomy increases its potential legal exposure relative to a narrower simulation-only vendor. SR014, SR026, SR034
CR014 No reviewed public source identified a current AV-incident lawsuit, consent order, or unresolved NHTSA enforcement action specifically naming Applied Intuition. SR018, SR014
CR015 Waymo v. Uber remains a durable example that autonomy software disputes can escalate into nine-figure trade-secret settlements and strategic disruption. SR019, SR022
CR016 Export-control risk for Applied is ongoing because every new foreign customer, transfer scenario, or defense use case can trigger fresh classification, licensing, and end-use analysis. SR010, SR011, SR012
CR017 RAND's safety work implies that statistically proving AV safety superiority can require from roughly 100 million to 100 billion miles, making the simulation-to-real gap structurally hard to close. SR020, SR021
CR018 NHTSA's ADS research and automated-vehicle safety pages still frame autonomy as an ongoing testing and oversight problem, not a solved validation problem. SR002, SR003
CR019 CARLA gives developers a credible free simulation baseline with APIs and community tooling, creating price pressure on any vendor whose value proposition is seen as simulation alone. SR030, SR023
CR020 Applied's likely moat versus CARLA is workflow integration, enterprise support, and defense-ready deployment context rather than an unassailable monopoly on simulation itself. SR026, SR030, SR031
CR021 Key-person risk remains meaningful because Applied is still publicly identified with CEO Qasar Younis while the company remains private and founder-shaped. SR034, SR036
CR022 Applied's hiring posture and the broader AV industry backdrop indicate that competition for autonomy, robotics, and AI engineering talent remains intense in 2026. SR022, SR029
CR023 Defense work adds a distinct hiring and collaboration constraint because export-controlled and procurement-sensitive programs narrow how talent and information can be deployed across projects. SR011, SR012, SR026
CR024 Embark is a direct cautionary analog showing that an autonomy company can build customer proof and still fail before reaching sustainable commercial scale. SR024, SR025, SR027
CR025 IEEE Spectrum's 2025 AV coverage presents a sector still wrestling with commercialization, safety incidents, and uneven business outcomes rather than a settled winner set. SR022, SR023
CR026 Autonomy programs at OEMs and trucking fleets can take years to move from validation into production, creating long payback cycles for infrastructure vendors like Applied. SR004, SR020
CR027 Applied's 17-of-top-20 OEM breadth does not disclose revenue concentration, contract duration, or production attachment, so concentration risk remains unresolved despite strong logo coverage. SR028, SR034
CR028 Revenue opacity is a core risk because Applied has public valuation and profitability narratives but no public revenue, ARR, or margin disclosure to verify sustainability. SR033, SR034, SR035
CR029 The $15 billion valuation raises the bar for future execution because later financing or IPO outcomes will be measured against a mark set without public operating data. SR034, SR035
CR030 Government and defense exposure diversifies end markets but adds budget-cycle, procurement, and program-priority risk that ordinary enterprise software vendors do not face. SR026, SR031, SR032
CR031 Geopolitical shocks can affect the same OEM and defense ecosystems that Applied sells into, amplifying supply-chain and exportability risk around major vehicle programs. SR021, SR031
CR032 Argo AI's 2022 shutdown demonstrates that even OEM-backed AV programs can be wound down before economics become durable. SR022, SR023
CR033 Embark raised substantial public capital yet still sold its assets to Applied, showing that autonomous-trucking timelines can outlast available financing. SR024, SR025, SR027
CR034 Motional's repeated restructuring reinforces that large sponsors do not automatically convert AV technical capability into durable commercial scale. SR022, SR023
CR035 TuSimple illustrates that autonomous-trucking strategies can also unravel through governance and geopolitical stress, not only through technology underperformance. SR022, SR023
CR036 Zoox and the Levandowski/Uber episode show two additional sector outcomes: strategic absorption instead of stand-alone success, and expensive IP fallout when talent mobility crosses legal boundaries. SR019, SR022, SR023
CR037 The worst-case regulatory scenario for Applied is a slow accretion of reporting, type-approval, export-control, and procurement obligations that raise cost without creating demand mandates. SR015, SR016, SR017
CR038 Applied does have meaningful mitigants, including broad OEM reach, defense traction, and a workflow-level product story that is less capital-intensive than operating AV fleets directly. SR026, SR028, SR031, SR032
CR039 Buying Embark's assets gave Applied trucking data, talent, and customer context without forcing it to inherit Embark's public-market fleet burn as a stand-alone operator. SR024, SR027, SR028
CR040 Export and defense compliance risk is manageable only if Applied maintains product classification, customer screening, controlled-sharing, and procurement discipline that are not publicly detailed today. SR010, SR011, SR012, SR013
CR041 Residual risk stays high because litigation history, ITAR classification status, DoD contract vehicle details, and customer-economics disclosure all remain thin in public evidence. SR026, SR033, SR034
CR042 Monitorable thesis-break signals would include a public AV liability suit naming Applied, major regulatory hardening without corresponding demand creation, or a down-round against the $15 billion narrative. SR014, SR015, SR034, SR035
CR043 Compared with robotaxi and self-driving trucking operators, Applied has lower direct on-road incident exposure but still faces indirect liability and demand risk because customer deployments drive tool spend. SR014, SR020, SR022
CR044 Applied's overall risk profile is better than failed pure-play fleet operators on capital intensity, but still materially riskier than generic enterprise software because autonomy regulation, safety, and defense compliance are core to the product. SR020, SR022, SR033, SR034
CV001 Applied Intuition announced a Series D round of 175 million dollars at a 3.6 billion dollar valuation. SV001, SV015, SV016
CV002 Applied Intuition disclosed an October 2024 financing of 250 million dollars at a 6 billion dollar valuation. SV002, SV010, SV016
CV003 Applied Intuition was reported at a 15 billion dollar valuation in its latest Series F financing with a 250 million dollar raise. SV003, SV006, SV007, SV008, SV009
CV004 BlackRock and Kleiner Perkins were associated with the latest round and reinforce an infrastructure-oriented investor base. SV003, SV005, SV007, SV009
CV005 Applied paired the latest financing with an OpenAI strategic partnership and a physical AI positioning narrative. SV003, SV005, SV006
CV006 Management said the company was profitable and growing triple digits year over year around the latest financing. SV003, SV006, SV009
CV007 Public detail for Applieds pre-Series D price history is sparse relative to its later rounds and secondary commentary. SV004, SV015, SV016, SV018, SV019
CV008 Applieds public valuation stepped up by roughly 4.2 times from 3.6 billion dollars at Series D to 15 billion dollars at the latest round. SV001, SV003, SV006, SV009
CV009 The move from 6 billion dollars in late 2024 to 15 billion dollars in the latest round implies a 2.5 times step-up in about one year. SV002, SV003, SV006, SV009
CV010 Analyst market reports place the pure autonomous vehicle simulation market in the low single digit billions rather than the tens of billions today. SV011, SV012, SV013
CV011 Analyst reports place the broader ADAS and autonomy software opportunity materially above the pure simulation segment. SV014, SV013
CV012 Applieds current company materials show the product story extending beyond simulation into vehicle intelligence and defense applications. SV032, SV033
CV013 The latest private valuation therefore appears to price a broader autonomy software and infrastructure thesis rather than a simulation-only business. SV003, SV032, SV033, SV011, SV014
CV014 Mobileye is the most relevant public comparable because it is a scaled autonomy software and ADAS platform with public-market price discovery. SV026, SV027
CV015 Mobileyes public valuation has been in the mid-teens billions, making Applieds 15 billion dollar mark comparable on headline enterprise value. SV026, SV027, SV017
CV016 Mobileyes disclosed revenue base is much larger than any public revenue disclosure available for Applied Intuition. SV026, SV027
CV017 Auroras public-market valuation remains far below 15 billion dollars despite its public filings and commercialization progress, underscoring market skepticism toward AV platforms. SV023, SV024, SV025
CV018 Scale AIs private valuation in the mid-teens billions shows that investors will pay Applied-like prices for AI infrastructure assets when strategic scarcity is believed. SV028, SV029, SV017
CV019 Waymo is commonly valued above Applied but is not a clean comp because it sits inside Alphabet and focuses on robotaxi deployment rather than software licensing. SV035, SV017, SV034
CV020 Waabi and Wayve represent earlier or narrower private autonomous software benchmarks than Applieds current pricing. SV030, SV031, SV017
CV021 Palantir shows that defense-linked AI platforms can sustain premium multiples when the market trusts execution and revenue quality. SV017, SV034
CV022 Embarks collapse and wind-down provide a stark adverse precedent for autonomy-sector valuation discipline. SV020, SV021, SV022
CV023 Applieds comparable set remains imperfect because public peers mix software, hardware, robotaxi, and defense exposure. SV017, SV023, SV026, SV028, SV030, SV031, SV035
CV024 Public sources do not disclose Applieds revenue, ARR, or gross margin, so a direct trailing revenue multiple cannot be computed from public data. SV003, SV015, SV016
CV025 A 6 billion dollar valuation would align with roughly 150 million to 250 million dollars of ARR at premium software multiples. SV002, SV011, SV014, SV015
CV026 A 15 billion dollar valuation would align with roughly 500 million to 750 million dollars of ARR at a 20 to 30 times multiple. SV003, SV014, SV017
CV027 If Applieds ARR is below 200 million dollars, the latest round implies a multiple of roughly 75 times ARR or higher. SV003, SV014, SV015
CV028 The profitability claim suggests better economic quality than is typical for venture-backed autonomy startups. SV003, SV006, SV009
CV029 Defense-linked revenue could justify a premium multiple because it diversifies demand and looks more like strategic infrastructure than a pure auto tool. SV033, SV005, SV034
CV030 The OpenAI partnership strengthens narrative premium but does not by itself prove monetized revenue today. SV003, SV005, SV006
CV031 BlackRock participation reinforces an infrastructure-style valuation framing rather than a conventional auto-supplier multiple. SV003, SV005, SV007, SV009
CV032 OEM tools licenses are the most plausible core revenue driver in Applieds current business model. SV032, SV015, SV016
CV033 Defense programs, vehicle intelligence licensing, and simulation usage are the most plausible upside drivers beyond core tools. SV033, SV032, SV005
CV034 The bull case requires Applied to surpass roughly 500 million dollars of ARR and prove platform-standard status across OEM and defense programs. SV003, SV014, SV017, SV033
CV035 The base case assumes about 250 million to 400 million dollars of ARR and sustained high growth, which still leaves the valuation somewhat stretched. SV014, SV015, SV016
CV036 The bear case assumes sub-200 million dollar revenue, slower OEM adoption, and multiple compression to 10 to 15 times. SV015, SV020, SV021, SV022, SV034
CV037 Probability-weighted scenario analysis clusters nearer low double digit billions than a clearly discounted entry price. SV015, SV016, SV017
CV038 Fresh 2026 revenue disclosure, larger defense awards, and deeper product standardization would be the most important bull-case validation signals. SV018, SV019, SV033, SV034
CV039 OEM budget slowing, simulation commoditization, or weak monetization outside core tools would be the clearest downside signals. SV011, SV012, SV022, SV034
CV040 Applied appears higher quality than many autonomy startups because it is profitable, multi-product, and backed by blue-chip investors. SV003, SV005, SV032, SV033
CV041 Public evidence is not strong enough to prove that a 15 billion dollar valuation is fair on fundamentals today. SV015, SV016, SV017, SV024, SV027
CV042 The most defensible public-data valuation stance is stretched rather than attractive or clearly fair. SV015, SV016, SV017, SV024, SV027
CV043 The bull thesis is that Applied becomes category-defining physical AI infrastructure for automotive and defense. SV003, SV005, SV033
CV044 The anti-thesis is that private-market AI exuberance and opaque financial disclosure are masking overvaluation. SV015, SV020, SV021, SV022, SV034
CV045 The right public-evidence recommendation is track rather than immediate full-conviction entry. SV015, SV016, SV017, SV034
CV046 The most decision-useful diligence asks are ARR, gross margin, segment mix, retention, and defense contract scale. SV015, SV016, SV018, SV019
CV047 Valuation confidence should remain medium because the price is explicit but the revenue denominator is still private. SV015, SV016, SV018, SV019
CV048 Earlier rounds and secondary marks remain too opaque to reconstruct a clean fully diluted share-price history from public evidence alone. SV004, SV015, SV016, SV018, SV019
来源
编号出版方标题引文
SO001 Applied Intuition Applied Intuition – Physical AI that moves the world
SO002 Applied Intuition About Applied Intuition
SO003 Applied Intuition Applied Intuition Products
SO004 Applied Intuition Applied Intuition Careers
SO005 Applied Intuition Applied Intuition Customers
SO006 Applied Intuition Applied Intuition Defense
SO007 Applied Intuition Series F announcement
SO008 Applied Intuition Series E announcement
SO009 Applied Intuition Series D announcement
SO010 Applied Intuition Series C press release
SO011 Applied Intuition Series B press release
SO012 Applied Intuition Embark acquisition press release
SO013 Applied Intuition OpenAI strategic partnership announcement
SO014 Applied Intuition U.S. Army customer post
SO015 Applied Intuition PACCAR partnership announcement
SO016 Applied Intuition Volkswagen partnership announcement
SO017 Applied Intuition Toyota partnership announcement
SO018 Applied Intuition GM partnership announcement
SO019 Applied Intuition Hyundai partnership announcement
SO020 Business Wire Applied Intuition Raises $250 Million Series F
SO021 Business Wire Applied Intuition Named to CNBC Disruptor 50
SO022 PACCAR PACCAR selects Applied Intuition
SO023 Volvo Group Volvo Group and Applied Intuition partnership announcement
SO024 CB Insights Applied Intuition company profile
SO025 PitchBook Applied Intuition company profile
SO026 Kleiner Perkins Applied Intuition perspective
SO027 CNBC Applied Intuition reaches $15 billion valuation
SO028 electrive Applied Intuition raises $250 million in Series F funding
SO029 Breaking Defense Applied Intuition defense coverage
SO030 U.S. Securities and Exchange Commission Embark Technology, Inc. DEFM14A filing index
SM001 MarketsandMarkets Autonomous Driving Simulation Market
SM002 MarketsandMarkets Autonomous Driving Software Market
SM003 Grand View Research Autonomous Vehicle Simulation Market
SM004 Allied Market Research Autonomous Vehicle Simulation Market
SM005 Mordor Intelligence Autonomous Vehicles Simulation Market
SM006 Mordor Intelligence ADAS Market
SM007 Business Research Insights Autonomous Vehicle Simulation Market
SM008 RAND Driving to Safety: How Many Miles of Driving Would It Take to Demonstrate Autonomous Vehicle Reliability?
SM009 RAND Autonomous Vehicle Safety and Policy Research
SM010 National Highway Traffic Safety Administration Automated Vehicles
SM011 National Highway Traffic Safety Administration Automated Vehicles for Safety
SM012 National Highway Traffic Safety Administration Federal Automated Vehicles Policy
SM013 National Highway Traffic Safety Administration Automated Driving Systems Research
SM014 Federal Register Automated Driving Systems — Federal Register Notice
SM015 GovTrack H.R. 3935
SM016 Congressional Research Service Autonomous Vehicles and Federal Policy
SM017 Regulations.gov Autonomous Vehicle Rulemaking Search Results
SM018 IEEE Spectrum Autonomous Vehicles
SM019 IEEE Spectrum Self-Driving Cars 2025
SM020 USASpending Autonomy-Related Defense Spending Search
SM021 Applied Intuition Applied Intuition
SM022 Applied Intuition Applied Intuition Defense
SM023 Applied Intuition Applied Intuition Products
SM024 Applied Intuition Applied Intuition and the U.S. Army
SM025 Aurora Innovation About Aurora
SM026 Waabi Waabi
SM027 Wayve About Wayve
SP001 dSPACE dSPACE Home
SP002 dSPACE dSPACE Press Releases
SP003 Cognata Cognata
SP004 ANSYS Autonomous Vehicle Simulation
SP005 IPG Automotive CarMaker
SP006 Metamoto Metamoto
SP007 VectorCAST VectorCAST
SP008 Scale AI Scale AI Company
SP009 GitHub CARLA Simulator Repository
SP010 CARLA CARLA Documentation
SP011 Waymo Waymo Open Dataset
SP012 Wayve About Wayve
SP013 Mobileye About Mobileye
SP014 Mobileye Mobileye News
SP015 Aurora About Aurora
SP016 Aurora Aurora Driver
SP017 Waabi Waabi
SP018 TechRadar Best Autonomous Vehicles
SP019 Applied Intuition Applied Intuition
SP020 Applied Intuition Applied Intuition Products
SP021 Applied Intuition Applied Intuition Physical AI
SP022 Applied Intuition Applied Intuition Defense
SP023 Applied Intuition Series F Announcement
SP024 MarketsandMarkets Autonomous Driving Simulation Market
SP025 IEEE Spectrum Autonomous Vehicles
SP026 IEEE Spectrum Self-Driving Cars 2025
SI001 Securities and Exchange Commission EDGAR Form D search results for 'applied intuition' Search results show Form D entries associated with Applied Intuition naming variants.
SI002 Securities and Exchange Commission EDGAR Form D search results for 'Applied Intuition Inc' Search results under the full corporate-name variant also return Applied Intuition Form D filings.
SI003 Securities and Exchange Commission Embark Technology EDGAR company page for DEFM14A materials
SI004 Securities and Exchange Commission Embark Technology DEFM14A filing index DEFM14A index for Embark proxy materials tied to the Applied Intuition transaction.
SI005 Embark Technology Investor Relations Embark Trucks investor relations site Investor-relations materials document the failed stand-alone path for Embark.
SI006 PACCAR PACCAR and Applied Intuition collaboration announcement
SI007 Bloomberg Applied Intuition reported at $15 billion valuation
SI008 Financial Times Applied Intuition Series F financing report
SI009 The Wall Street Journal Applied Intuition raises $250 million in funding round
SI010 Applied Intuition Series F announcement
SI011 Applied Intuition Series E announcement Applied Intuition is profitable and growing at a sustainable triple-digit percentage year-over-year.
SI012 Applied Intuition Series D announcement
SI013 Applied Intuition Series C press release
SI014 Applied Intuition Series B press release
SI015 Applied Intuition Embark acquisition press release
SI016 Applied Intuition Embark transaction blog post
SI017 Applied Intuition Embark acquisition completion blog post
SI018 Applied Intuition Applied Intuition home page
SI019 Applied Intuition Applied Intuition about page
SI020 Applied Intuition Applied Intuition careers page
SI021 BusinessWire Applied Intuition raises $250 million financing
SI022 CB Insights Applied Intuition company profile
SI023 PitchBook Applied Intuition company profile
SI024 electrive Applied Intuition raises $250 million in late-2024 round
SI025 EE Times Applied Intuition raises $250 million financing
SI026 The Decoder Applied Intuition reported at $15 billion valuation
SI027 CNBC Applied Intuition valued at $15 billion
SI028 Defense News Applied Intuition defense and autonomy coverage
SI029 Breaking Defense Applied Intuition defense and physical AI coverage
SI030 Kleiner Perkins Applied Intuition perspective
SI031 Volvo Group Volvo Group and Applied Intuition partnership
SI032 Aurora Aurora company about page
SI033 Waabi Waabi company page
SI034 Waymo Waymo Open
SI035 Scale AI Scale AI company page
SE001 Applied Intuition Applied Intuition | Physical AI that moves the world™
SE002 Applied Intuition Products | AI platforms for vehicle autonomy | Applied Intuition
SE003 Applied Intuition Autonomous vehicle platform | ADAS & AV | Applied Intuition
SE004 Applied Intuition Software defined vehicle OS | Applied Intuition
SE005 Applied Intuition Physical AI | Simulation, verification & validation | Applied Intuition
SE006 Applied Intuition ADAS & autonomous vehicle simulation | Applied Intuition
SE007 Applied Intuition Autonomous trucking & driverless trucks | Applied Intuition
SE008 Applied Intuition Autonomous precision agriculture technologies | Applied Intuition
SE009 Applied Intuition Construction robotics & automation | Applied Intuition
SE010 Applied Intuition Autonomous mining & fleet management system | Applied Intuition
SE011 Applied Intuition Automated defense tech & AI defense company | Applied Intuition
SE012 Applied Intuition Army ISV made autonomous in 10 days | Applied Intuition
SE013 Applied Intuition Customer stories & case studies | Applied Intuition
SE014 Applied Intuition Autonomy & robotics research | Applied Intuition
SE015 Applied Intuition Series F funding drives $15B valuation | Applied Intuition
SE016 GitHub GitHub - carla-simulator/carla: Open-source simulator for autonomous driving research.
SE017 CARLA CARLA Simulator
SE018 Ansys ADAS & Autonomous Vehicle Simulation Software | Ansys
SE019 IPG Automotive CarMaker
SE020 dSPACE dSPACE Home
SE021 arXiv One Thousand and One Hours: Self-driving Motion Prediction Dataset
SE022 Waymo About – Waymo Open Dataset
SE023 Mobileye About Mobileye | Our Vision, History, and Milestones
SE024 SambaNova Resources | Blog
SE025 AWS Marketplace AWS Marketplace: 400
SE026 Scale AI Scale AI - Data Platform for AI
SU001 Applied Intuition Applied Intuition customers page
SU002 Applied Intuition Applied Intuition case studies
SU003 Applied Intuition Embark acquisition press release Embark had 17 of the top 20 global automotive OEMs as customers.
SU004 Applied Intuition Series E announcement The company is profitable and growing at a sustainable triple-digit percentage year-over-year.
SU005 Applied Intuition Toyota partnership announcement
SU006 Applied Intuition Volkswagen partnership announcement
SU007 Applied Intuition GM partnership announcement
SU008 Applied Intuition Hyundai partnership announcement
SU009 Applied Intuition PACCAR partnership announcement
SU010 PACCAR PACCAR selects Applied Intuition PACCAR has selected Applied Intuition to support autonomous truck development.
SU011 Volvo Group Volvo Group and Applied Intuition partnership announcement
SU012 Applied Intuition U.S. Army customer post
SU013 U.S. Army Army partners with Applied Intuition
SU014 U.S. Department of Defense Army secretary talks Applied Intuition They had made our vehicles fully autonomous. Seven weeks later, our soldiers are actually testing that equipment.
SU015 Applied Intuition Applied Intuition defense page
SU016 Applied Intuition Applied Intuition defense-tech page They had made our vehicles fully autonomous. Seven weeks later, our soldiers are actually testing that equipment.
SU017 Applied Intuition Applied Intuition autonomous trucking page Through our strategic partnership with Applied Intuition, we have made significant progress in the development of autonomous driving technology to address the challenges that the Japanese logistics industry is currently facing.
SU018 Applied Intuition Army autonomous ISV post
SU019 Applied Intuition Applied Intuition autonomous vehicles page
SU020 Applied Intuition Applied Intuition agricultural automation page
SU021 Applied Intuition Applied Intuition autonomous construction equipment page
SU022 Applied Intuition Applied Intuition mining fleet management system page By integrating Komatsu's deep industry experience with cutting-edge vehicle software from Applied Intuition, we are accelerating progress toward safer, more connected, and environmentally responsible mining operations.
SU023 Applied Intuition Series F announcement
SU024 The Robot Report Applied Intuition raises $250M series-f at $6B valuation
SU025 PitchBook Applied Intuition company profile
SU026 CB Insights Applied Intuition company profile
SU027 Breaking Defense Applied Intuition defense coverage
SU028 Defense News Applied Intuition defense and autonomy coverage
SU029 CNBC Applied Intuition reaches $15 billion valuation
SU030 Business Wire Applied Intuition raises $250 million Series F
SU031 Applied Intuition OpenAI strategic partnership announcement
SU032 Applied Intuition Nuro customer post
SU033 Applied Intuition U.S. Army autonomous vehicles post
SU034 OpenAI OpenAI and Applied Intuition news page
SU035 Automotive News Applied Intuition automotive software startup funding article
SU036 Business Wire Applied Intuition enables PACCAR drive development
SR001 National Highway Traffic Safety Administration Federal Automated Vehicles Policy
SR002 National Highway Traffic Safety Administration Automated Driving Systems Research
SR003 National Highway Traffic Safety Administration Automated Vehicles
SR004 Congressional Research Service Autonomous Vehicles and Federal Policy
SR005 GovTrack H.R. 3935
SR006 Regulations.gov Autonomous Vehicle Rulemaking Search Results
SR007 Federal Motor Carrier Safety Administration Safe Integration of Automated Driving Systems (ADS)-Equipped Commercial Motor Vehicles (CMVs)
SR008 National Conference of State Legislatures Autonomous Vehicles Legislation Database
SR009 National Institute of Standards and Technology AI Risk Management Framework
SR010 Bureau of Industry and Security Licensing | Bureau of Industry and Security
SR011 International Trade Administration U.S. Export Controls
SR012 Acquisition.GOV DFARS | Acquisition.GOV
SR013 Acquisition.GOV FAR | Acquisition.GOV
SR014 Greenberg Traurig LLP Self-Driving Vehicles: Liability Assignment in Crashes and Violations
SR015 Sidley Austin LLP — Environmental Health & Safety Brief NHTSA Proposes Amending Federal Crash Avoidance Standards for Autonomous Vehicles
SR016 Varnum LLP NHTSA Adds Three Rulemakings That Will Impact Automated Vehicles
SR017 Mayer Brown LLP DOT and NHTSA Announce Autonomous Vehicle Framework
SR018 Husch Blackwell LLP NHTSA Closes Waymo Investigation — Key Takeaways for the AV Industry
SR019 Forensic Analysis Group (Forensisgroup) Waymo LLC v. Uber Technologies Inc. — Trade Secret Theft and Self-Driving Car Technology Litigation
SR020 RAND Driving to Safety: How Many Miles of Driving Would It Take to Demonstrate Autonomous Vehicle Reliability?
SR021 RAND Autonomous Vehicle Safety and Policy Research
SR022 IEEE Spectrum Self-Driving Cars 2025
SR023 IEEE Spectrum Autonomous Vehicles
SR024 Securities and Exchange Commission Embark Technology DEFM14A filing index
SR025 Embark Technology Investor Relations Embark Trucks investor relations site
SR026 Applied Intuition Applied Intuition defense page
SR027 Applied Intuition Embark acquisition post
SR028 Applied Intuition Embark acquisition press release
SR029 Applied Intuition Applied Intuition careers page
SR030 CARLA Simulator CARLA GitHub repository
SR031 Breaking Defense Applied Intuition defense coverage
SR032 Defense News Applied Intuition defense and autonomy coverage
SR033 Applied Intuition Series E announcement
SR034 Applied Intuition Series F announcement
SR035 CNBC Applied Intuition reaches $15 billion valuation
SR036 Applied Intuition About Applied Intuition
SV001 Applied Intuition Series D announcement
SV002 Applied Intuition Series E announcement
SV003 Applied Intuition Series F announcement
SV004 Applied Intuition Series C press release
SV005 Kleiner Perkins Applied Intuition perspective
SV006 The Decoder Applied Intuition 15 billion valuation article
SV007 Bloomberg Applied Intuition raises funding at 15 billion valuation
SV008 Reuters Applied Intuition 15 billion valuation report
SV009 The Robot Report Applied Intuition closes Series F at 15 billion valuation
SV010 Axios Applied Intuition funding coverage
SV011 Grand View Research Autonomous vehicle simulation market
SV012 Business Research Insights Autonomous vehicle simulation market report
SV013 MarketsandMarkets Autonomous driving simulation market
SV014 MarketsandMarkets Autonomous driving software market
SV015 PitchBook Applied Intuition company profile
SV016 CB Insights Applied Intuition company profile
SV017 CB Insights Autonomous vehicle funding unicorn tracker
SV018 Securities and Exchange Commission Applied Intuition Form D search
SV019 Securities and Exchange Commission Applied Intuition Inc Form D search
SV020 Securities and Exchange Commission Embark DEFM14A browse page
SV021 Securities and Exchange Commission Embark DEFM14A filing index
SV022 Business Wire Embark Trucks announces wind down of operations
SV023 Securities and Exchange Commission Aurora 10-K browse page
SV024 Securities and Exchange Commission Aurora annual report
SV025 Aurora Innovation Investor relations
SV026 Mobileye About Mobileye
SV027 Securities and Exchange Commission Mobileye Global 20-F search
SV028 Scale AI About Scale AI
SV029 Scale AI Scale AI Series F announcement
SV030 Waabi Waabi homepage
SV031 Wayve About Wayve
SV032 Applied Intuition About Applied Intuition
SV033 Applied Intuition Applied Intuition Defense
SV034 CNBC Autonomous vehicle industry risks and regulation
SV035 Waymo Waymo Open overview
SV036 Second Measure Bloomberg Second Measure