Stardust Intelligence

1.1 Identity, product, and operating model

Available reporting identifies Stardust Intelligence as Astribot, a Shenzhen-based company founded in 2022. Its founder and CEO, Lai Jie, is described as having more than 17 years of AI and robotics research and design experience, including work at Tencent's Robotics Laboratory and Baidu's Xiaodu Robot team. The reported product strategy is vertically integrated: a base model, the Embodied OS, and a rope-driven body are developed as one stack. The rope-driven design places motors away from joints and uses tendon-like cables, while the DuoCore framework separates fast real-time response from slower planning. These descriptions are primarily company or company-linked claims rather than independently audited technical results.[CO001, CO002, CO003, CO004]

1.2 Leadership, financing, and governance visibility

The strongest current evidence concerns financing rather than governance. Two June 2026 reports state that the company completed three Series B financings within three months, raising more than RMB1 billion and reaching a valuation above RMB10 billion. The reported investors span state-linked funds, listed-company strategic investors, financial investors, and continuing shareholders. The same sources say proceeds are intended for model iteration, real-world data, and recruiting. No source in the assigned pool supplies a board roster, audited cap table, detailed governance structure, or a documented leadership bench beyond Lai Jie. Those omissions matter because the reported valuation is material while the available evidence does not establish revenue, margins, cash burn, or ownership concentration.[CO005, CO006, CO007, CO012]

1.3 Commercial milestones and unresolved diligence

The reported milestone sequence moves from founding to productization, claimed thousand-unit delivery activity at the end of 2025, a T-series launch in May 2026, and the June 2026 financing. One report describes a thousand-unit industrial and commercial service order with Thundersoft and a proposed application innovation center with Jiangdu Economic Development Zone; these are useful commercialization signals but do not disclose recognized revenue, delivered units, customer names, renewal rates, or unit economics. External sector sources show that humanoid robotics is moving toward production scale while also warning about reliability, cost, dexterity, and hype. For Stardust specifically, the assigned pool does not provide independent confirmation of its delivery count, manufacturing yield, actuator reliability, revenue, or board and executive coverage. Those are material refresh gaps rather than assumptions to fill.[CO004, CO008, CO009, CO010, CO011, CO013, CO014, CO015, CO016]

1.4 Exhibits

2.1 Market boundary and buying problem

Stardust Intelligence participates in the embodied-AI and humanoid-robot market, but the investable opportunity should be bounded around physical systems that combine a robot body, onboard or connected intelligence, and an operating workflow. Included spend is the purchase or lease of the robot, deployment integration, safety and maintenance, teleoperation, model or software support, and training data generated in use. Excluded from the core boundary are general industrial automation that does not need a humanoid form, research demonstrations with no deployment budget, and broad AI software revenue that is not attached to physical operation. The closest substitutes are fixed automation, specialized mobile robots, human labor, and teleoperated equipment. Buyer value is therefore task-specific: a lower-priced body can expand pilots, but it does not by itself prove repeatable utilization, safe operation, or an economic replacement for an existing workflow.[CM001, CM002, CM003, CM004]

2.2 Multiple sizing lenses and commercialization signal

The evidence supports a layered sizing view rather than a single top-down TAM. A global shipment lens reported approximately 15,000 units expected in 2025, while another market estimate cited roughly 18,000 global shipments for that year; these are forecasts or estimates with different methodologies, not a reconciled installed base. A China production lens is more actionable for Stardust: TrendForce projects Chinese humanoid output growth of up to 94% in 2026 and nearly 80% combined share for Unitree and AgiBot, while the TechTimes account describes AgiBot reaching 10,000 produced units and Unitree reporting 5,500 humanoid units shipped in 2025. A company-level price lens is also visible: the newly reported T1 starting price below RMB90,000 may lower the pilot barrier, but it is not evidence of realized revenue, gross margin, or total serviceable demand. These lenses indicate a fast-forming supply market, not a validated broad market for autonomous general-purpose labor.[CM005, CM006, CM007, CM008, CM009]

2.3 Buyers, adoption drivers, and constraints

Near-term adoption is most plausible where the buyer can define a repeatable task and tolerate human oversight: factories, logistics sites, retail or hospitality venues, education and research labs, and public-service pilots. The budget owner may be operations, automation engineering, innovation, research, or workforce-training leadership rather than a general IT function. Drivers include labor scarcity, a lower hardware entry price, access to real-world training data, and the possibility of integrating a robot into an existing operating system or channel. Constraints are equally material. A market account cites a Morgan Stanley survey in which only 23% of prospective industrial buyers reported satisfaction with currently available products; Futurism quotes warnings that inference, dexterity, reliability, and cost still confine initial use cases to structured factories and warehouses. Battery endurance of two to three hours, uncertain sustained utilization, cybersecurity or foreign-adversary policy concerns, and the engineering cost of adapting workflows can delay adoption even when a pilot is technically successful. For valuation, the key proof point is not shipment count but paid, repeatable utilization with disclosed deployment economics.[CM010, CM011, CM012, CM013, CM014]

2.4 Sizing and adoption gaps

The retained evidence does not isolate Stardust Intelligence's SAM or SOM by geography, workflow, or paying customer. Reported orders and deployments are largely company or investor communications, and the pool does not provide audited revenue, realized average selling price, service attach rate, utilization, failure rates, or customer retention. The market estimates also mix shipments, production, and projected output, so they cannot be converted into a common revenue pool without a consistent unit, period, geography, and price basis. The next diligence step is to obtain a deployment cohort by use case with delivered units, paid status, uptime, intervention hours, gross margin after service, and repeat-order behavior, then reconcile it to a bottom-up list of target sites and budget owners. Until that evidence exists, the prudent market conclusion is that the category is expanding rapidly but the company-specific serviceable opportunity remains evidence-constrained.[CM015, CM016]

2.5 Exhibits

3.1 Competitive landscape and scale signals

The relevant landscape includes direct humanoid peers, vertically integrated incumbents, adjacent component and platform providers, and non-humanoid substitutes for structured work. Public reporting places Unitree and AGIBOT at the leading edge of Chinese shipment and production scale: Omdia estimates cited by CNBC put expected 2025 global shipments at 15,000 units, with Unitree second to AGIBOT, while TrendForce projects the two companies together could represent nearly 80% of Chinese shipments in 2026. Galbot's company-issued release claims more than $300 million of new funding, $800 million total funding, a $3 billion valuation, thousands of industrial orders, and operations in more than 30 cities; those claims are useful competitive signals but are not independent verification. Stardust therefore enters a field where scale, ecosystem access, and deployment evidence may matter as much as mechanical differentiation.[CP001, CP002, CP003, CP004]

3.2 Capability, distribution, and trust comparison

The strongest documented peer advantages are differentiated rather than uniform. Unitree has a low-price product ladder and visible capital-market access; AGIBOT has a reported 10,000-unit production milestone and broad deployment claims; Galbot describes a full-stack model, industrial partners, retail, logistics, and hospital use cases. Stardust has a distinct rope-driven architecture and a documented three-party development-platform order with Baidu Intelligent Cloud and Jishu Iteration, which could support data collection and developer distribution. However, the evidence does not establish that this platform order equals recurring robot revenue, nor that rope-driven actuation has superior lifetime reliability or cost. Security and standards posture is also a buying criterion: Chinese humanoid vendors face scrutiny around safety, data, and compliance, while emerging standards may reduce interoperability uncertainty without proving any vendor's compliance.[CP005, CP006, CP007, CP008]

3.3 Pricing, switching costs, and moat durability

Public pricing evidence is concentrated in Unitree rather than Stardust or most peers. CNBC reports a $5,900 starting price for Unitree R1 and a $16,000 starting price for G1, while contrasting Tesla Optimus at an expected price of at least $20,000. These are list or reported starting prices, not realized enterprise contract prices, service costs, gross margins, or total cost of ownership. Galbot and AGIBOT emphasize deployment scale and partnerships rather than public price cards. Switching costs are likely to arise from data pipelines, task models, operator training, safety validation, and component supply, but the assigned evidence does not quantify customer lock-in, multi-homing, or partner exclusivity. Stardust's Baidu/Jishu platform relationship may lower adoption friction for educational and development use, yet customer retention and recurring software economics remain open questions.[CP009, CP010, CP011]

3.4 Displacement risk and diligence judgment

Competitive displacement risk is material because production is scaling faster than independently verified end-use demand. Futurism cites warnings that humanoid startups face fundamental challenges in inference, dexterity, reliability, and cost, limiting initial use cases to structured factories and warehouses; it also reports concerns about an investment bubble and similar products crowding out research. Robotics and Automation News describes China's 2026 safety-and-ethics standards framework, but also notes that tactile sensing remains a bottleneck. This creates a two-sided risk for Stardust: a standards-led market can accelerate adoption, while better-funded peers may commoditize hardware before Stardust proves a durable actuator, data, or distribution moat. The next diligence priority is comparative evidence on delivered units, uptime, service economics, and named customer expansion rather than additional valuation headlines.[CP012, CP013, CP014]

3.5 Exhibits

4.1 Revenue model and monetization

Stardust Intelligence's public materials in the assigned evidence pool do not disclose a price list, realized revenue, customer contracts, revenue mix, or accounting policy. The available evidence therefore supports a mechanism-level view rather than a company revenue estimate. Humanoid robotics monetization can begin with hardware sales or pilots, then expand through deployment services, software or model licensing, maintenance, and partnerships; however, the mix for Stardust Intelligence is an open question. Unitree is a useful market reference, not a proxy for Stardust: CNBC reported that Unitree had revenues exceeding 1 billion yuan and had been profitable since 2020, while also reporting a starting price of $5,900 for its R1 and a $16,000 starting price for its G1. Those are list or reported company figures from a different issuer, so they should not be converted into Stardust revenue or gross margin. The underwriting conclusion is that a credible revenue bridge requires delivered units, realized price, deployment revenue, repeat service revenue, and customer-level concentration.[CI001, CI002, CI003, CI004]

4.2 Pricing, GTM, and sales efficiency

Public sector evidence points to a transition from demonstrations toward structured industrial deployments, but it does not establish Stardust's sales cycle, CAC, payback, discounting, or channel economics. TrendForce reported that Chinese vendors were clarifying commercial use cases and that order-driven manufacturing and dedicated supply arrangements were supporting scale; its report also described Unitree's humanoid revenue as more than 51% of total revenue in 2025 and combined gross margin across its segments at 60%. The figures are sector and competitor evidence, not Stardust disclosures. CNBC's account of Unitree's low-price products suggests a possible hardware-led acquisition and data-collection strategy, while the Futurism analysis warns that reliability, inference, dexterity, and cost constrain early use cases to controlled factories and warehouses. For Stardust, this implies that sales efficiency should be measured by paid deployments that reach productive utilization, not by announcements, pilots, or headline orders.[CI005, CI006, CI007, CI008]

4.3 Cost structure and unit economics

No assigned source provides Stardust's bill of materials, labor content, warranty reserve, gross margin, installation cost, service burden, working capital, capex, or manufacturing yield. These omissions are material because a cable-driven humanoid can have a different actuator, maintenance, and reliability cost curve from a conventional design, yet the available sources do not quantify Stardust's economics. TrendForce reported that component localization and mass-production scale were improving the sector's cost base, while its discussion of capacity expansion and upstream components illustrates the capital commitment required for volume. The Futurism source provides adverse sector context: experts warned that startups remained a long way from generating profits and should pursue early revenue streams with a clear monetization model. Unit-economics underwriting should therefore keep all Stardust values null until source-backed disclosures exist and should request contribution margin by product, deployment labor, spares, warranty claims, utilization, and cash conversion.[CI009, CI010, CI011]

4.4 Capital adequacy and financial verdict

The assigned pool does not disclose Stardust's cash balance, burn rate, runway, debt, project-finance obligations, financing covenants, or planned use of funds. The prior report context identifies a valuation above RMB10 billion and three Series B rounds, but those facts are not current chapter evidence and are not sufficient to infer cash available or runway. Comparable financing activity demonstrates both access to capital and valuation risk: PR Newswire reported that Galbot completed a funding round exceeding $300 million and reached a $3 billion valuation, while CNBC reported Unitree's proposed valuation could reach $7 billion and that it was unclear how much the IPO would raise. Futurism's adverse evidence cautions that investment enthusiasm may outpace profitability. Stardust should be treated as financing-dependent until management supplies a cash bridge, monthly burn, committed orders, capex schedule, debt terms, and a next-round trigger. The financial verdict is research-more: commercial scale may be plausible, but revenue quality, margin path, and liquidity cannot yet support valuation underwriting.[CI012, CI013, CI014]

4.5 Exhibits

5.1 Product definition and operating context

Stardust Intelligence is publicly described in the assigned pool only through a NetEase article title as offering the rope-driven T1 AI robot, including food preparation and experiment-oriented use cases; the fetched page body is not available beyond its title/navigation shell, so those capabilities remain company-specific evidence gaps rather than verified operating claims. Sector evidence indicates that near-term humanoid value is concentrated in controlled automotive and logistics workflows rather than general household autonomy. IDTechEx describes early tasks as material handling, inspection support, intra-factory transport, and simple assembly assistance, with reliability, safety validation, and maintainability prioritized over general-purpose capability. This makes the practical product question less about demonstrations and more about repeatable task execution in a bounded customer workflow. China’s 2026 standards framework also spans complete systems, applications, safety, ethics, data, training, inference, and deployment, raising the diligence bar for any product claiming embodied intelligence. The current record does not establish Stardust’s buyer, production-versus-pilot status, software interface, or measurable customer outcome.[CE001, CE002, CE003, CE004]

5.2 Architecture, dependencies, and maturity

The only detailed mechanical benchmark in the pool is Unitree’s G1 page: a comparable humanoid platform lists 23 degrees of freedom, crossed-roller bearings, internal-rotor permanent-magnet synchronous motors, hollow electrical routing, dual encoders, local air cooling, depth camera plus 3D LiDAR, an eight-core CPU, Wi-Fi 6/Bluetooth 5.2, and about two hours of battery life. These fields illustrate the subsystem stack Stardust must document, but they do not establish that Stardust uses the same design. Independent market analysis identifies battery energy density, thermal management, precision screws and bearings, actuators, tactile sensing, and end-effectors as scaling bottlenecks. The standards source further identifies lifecycle data, model training, inference, deployment, application interfaces, trustworthiness, and evaluation as active standardization areas. Against that benchmark, the rope/cable-driven approach is potentially differentiated, but the assigned evidence contains no verified actuator topology, transmission lifetime, calibration process, software stack, API, sensor suite, MTBF, manufacturing yield, or field-service design for Stardust.[CE005, CE006, CE007, CE008, CE009]

5.3 Reliability, safety, and implementation risk

Product risk is elevated because humanoid systems combine high-energy mechanics, batteries, perception, control software, and human proximity. Unitree’s published page warns that the robot has a complex structure and powerful actuation, asks users to maintain a sufficient safety distance, and says some sample functions remain under development and testing; those statements are a useful benchmark for the type of qualification Stardust should publish, not evidence of Stardust’s own safety record. The new Chinese standards framework includes safety and ethics across the lifecycle, while Shenzhen’s insurance initiative shows that embodied-robot risk protection is becoming an infrastructure concern. Sector research says home use remains longer term because safety, affordability, and reliability barriers remain unresolved, and notes battery runtime and component supply-chain limits. The refresh pool contains no Stardust incident history, certification, security assessment, uptime data, deployment acceptance test, warranty terms, or developer signal. Those omissions materially limit underwriting of reliability and implementation cost.[CE010, CE011, CE012, CE013]

5.4 Exhibits

6.1 Buyer, user, and payer segments

The clearest target segment is industrial and logistics operations, where robots can move materials, transport turnover boxes, load and unload lines, and recover empty containers. A reported order with Xi'an Gong Intelligence describes a two-year, staged deployment across industry, manufacturing, warehousing, and logistics, with Xi'an Gong providing control-system technology and scenario integration. This is a workflow-level signal rather than proof of paid end-customer usage: the public item does not disclose price, contract value, delivery schedule by tranche, acceptance criteria, or the ultimate operating sites. A separate company-linked product article places Stardust in industrial logistics, commercial tourism, eldercare, research, and education, while describing T1 as aimed at household, commercial, research, entertainment, and industrial settings. Those broader segments should be treated as go-to-market targets, not equivalent evidence of mature revenue.[CU001, CU002, CU003, CU004]

6.2 Adoption trajectory and proof quality

The strongest observed trajectory is a move from demonstrations toward structured industrial workflows, not generalized household adoption. TrendForce reports that Chinese vendors are clarifying commercial use cases and scaling output, and says automotive manufacturing, consumer electronics, and logistics orders indicate a shift from trial adoption toward real demand across the sector. Independent commentary nevertheless describes current practical use as concentrated in structured manufacturing and logistics, with many systems still scripted or teleoperated and with a need to see robots working alongside humans in normal production sites before calling them ready for commercial scale-up. Stardust-specific materials claim thousand-unit-scale batch delivery and scenario ROI models, but provide no audited utilization, uptime, repeat-order, or customer outcome dataset. Therefore the adoption signal is promising but should not be converted into a customer count or recurring-revenue claim.[CU005, CU006, CU007, CU008]

6.3 Named proof, durability, and reference quality

Xi'an Gong Intelligence is the only named Stardust counterparty in the retained evidence. The reported arrangement combines Stardust's rope-driven robot and embodied-intelligence work with Xi'an Gong's control technology, and describes staged deployment into industrial and logistics scenarios. That makes it stronger than a logo or generic partnership mention, but the source is a short news item repeating a company announcement and does not identify the end-buyers, locations, deployed-versus-ordered units, or measured production outcomes. The T1 article further claims completed batch deliveries across four broad lanes and repeat purchase intent, yet those assertions are promotional and lack customer quotations or independently audited records. Retention metrics such as GRR, NRR, churn, renewal length, cohort survival, and support burden are not disclosed in the pool.[CU009, CU010, CU011, CU012]

6.4 Expansion path and concentration risk

A plausible expansion loop is to start with controlled industrial material handling, use deployment data to improve the robot and model, then extend into other structured commercial or service workflows. The reported order explicitly describes staged, multi-scenario deployment and continued data collection, while independent market evidence identifies manufacturing, warehousing, and logistics as the earliest adoption settings because their layouts and tasks are comparatively repeatable. That loop remains a hypothesis until Stardust discloses paid deployment cohorts, acceptance and renewal terms, utilization by site, unit economics, and repeat-order conversion. Concentration risk is also unresolved: one named integration counterparty and undisclosed end-customer identities make it impossible to assess whether demand is diversified or dependent on a single channel, partner, or anchor order. Procurement friction, safety acceptance, and serviceability may further slow expansion.[CU013, CU014, CU015, CU016]

6.5 Exhibits

7.1 Regulatory and geopolitical exposure

The clearest external risk is that foreign-made robotics are becoming a national-security policy category. The U.S. GUARD Act proposal would subject certain humanoid and quadruped robots, plus their communications systems, to national-security review and could place products on the FCC Covered List if they are judged an unacceptable risk (SR001, SR002). The proposal is not an enacted prohibition, so its direct legal effect on Stardust Intelligence is unresolved; however, it signals a pathway by which customer eligibility, import, research, and public-sector procurement could tighten. This is a jurisdictional and market-access risk rather than evidence that Stardust has been accused of wrongdoing. Investors should require an export-control map, target-country screening, and a contingency plan for restricted deployments.[CR001, CR002, CR003]

7.2 Cybersecurity, safety, and operational reliability

Connected humanoids combine physical actuation, sensors, software updates, and networked control, so a cyber incident can become a safety, service, and liability event. The GUARD Act announcement frames the policy concern as sensitive data exposure, persistent access, or remote disruption in connected robotics (SR002), while reporting on the sector notes that Unitree researchers disclosed a wormable Bluetooth flaw affecting robots and described full-device takeover; that example is not evidence of a Stardust vulnerability, but it demonstrates the class of failure investors must test (SR005). Public sources in this pool do not establish Stardust's penetration-test results, incident history, safety certification, mean time between failures, or patch-service commitments. Those unknowns leave residual exposure high until independent testing, update signing, offline recovery, and field-failure data are supplied.[CR004, CR005, CR006]

7.3 Demand quality and commercialization risk

Sector-level commercialization evidence is mixed. A KrASIA account reports investors describing embodied intelligence as commercially unclear, says many automotive deployments may be strategic tie-ups rather than sales, and quotes an investor questioning valuations before product-market fit (SR003). By contrast, TrendForce reports rapid Chinese output growth and says Unitree and AgiBot could account for nearly 80% of shipments, while CNBC reports strong sector investment and planned deployments for other startups (SR006, SR009). These are market context, not Stardust customer proof. The risk is that headline orders or demonstrations do not convert into paid, repeatable production deployments with acceptable service economics. Stardust has not publicly disclosed enough independently verified delivered-unit, utilization, renewal, gross-margin, or customer-concentration data in this pool to quantify that risk.[CR007, CR008, CR009]

7.4 Capital intensity, dependencies, and thesis-break triggers

Humanoid robotics requires coordinated hardware, software, manufacturing, and component supply, making execution sensitive to financing conditions and vendor continuity. TrendForce describes planned capacity expansion and upstream demand for joint modules, hands, and sensors, while CNBC reports that Chinese humanoid startups are rushing toward public markets amid rapidly rising sector investment (SR006, SR009). Galbot's own release says its latest round exceeded $300 million and assigned the company a $3 billion valuation, while SCMP reports that BYD and Xpeng are accelerating humanoid mass-production plans (SR007, SR008). Funding raised is not cash available, but these signals show intensifying competition for talent, suppliers, and commercialization milestones. The pool does not provide Stardust's cash balance, burn, debt, component single-source exposure, manufacturing yield, or warranty reserve. A thesis should therefore break if the company cannot fund its next production milestone, loses a critical supplier without a qualified substitute, or fails independent reliability and paid-deployment tests.[CR010, CR011, CR012, CR013, CR014, CR015]

7.5 Exhibits

8.1 Investment call and valuation discipline

The investable thesis is that Stardust can convert a differentiated cable-driven actuator, a vertically integrated model/OS/body stack, and a timely product launch into valuable industrial automation. The anti-thesis is that market growth and peer financing marks do not prove Stardust's utilization, reliability, gross margin, or cash conversion. IDTechEx forecasts an approximately US$29.5bn market by 2036, while Robotics & Automation News, reporting IDTechEx analysis, says average selling prices could fall from approximately $114,700 in 2024 to around $37,000 by 2030 and warns that deployment efficiency and task continuity determine realized economics. Goldman Sachs frames a much wider outcome, from at least US$6bn to as much as US$154bn by 2035E, explicitly conditional on overcoming design, use-case, technology, affordability, and acceptance hurdles. These are market scenarios, not Stardust forecasts. The reported RMB15bn mark should therefore be treated as an unverified financing reference, not intrinsic value. With no public audited revenue, margin, shipment, customer, cap-table, preference, or dilution schedule in this pool, the appropriate call is track/research-more; a buy recommendation would require false precision.[CV001, CV002, CV003, CV004, CV005, CV012, CV015]

8.2 Thesis, anti-thesis, and price-sensitive proof

The positive case has three evidence-backed legs. First, the market is moving toward structured industrial and logistics deployments, where clear task boundaries make ROI more testable. Second, peer production and financing activity indicate capital-market appetite: TrendForce reports that Unitree and AgiBot could account for nearly 80% of Chinese shipments in 2026, while Galbot's company-issued release states a $3 billion valuation after a funding round exceeding $300 million. Third, the Baidu Baike entry reports T1's 1.55-metre, 66-kilogram body, 23 degrees of freedom, a 5-kilogram single-arm payload, and a reported October financing mark of RMB15bn. The counterweight is material: those Stardust financing and product facts are not independently audited in the retained evidence, and the peer evidence is not a comparable valuation set. IDTechEx says favorable payback can be about six months only under high-utilization assumptions and about 15 months under medium utilization; CNBC reports the current market is only $2bn to $3bn according to a Barclays researcher and describes commercialization as still developing. The price call changes only when Stardust supplies cohort-level utilization, realized ASP, service costs, warranty/repair rates, and financing terms.[CV006, CV007, CV008, CV009, CV010, CV011, CV012, CV013]

8.3 Scenarios, comparables, and exit readiness

A scenario framework is more defensible than a point valuation. The bull case assumes T1 achieves repeatable structured-workflow deployments, high utilization, falling unit cost, and enough reliability to support renewals; the base case assumes pilots and selected industrial tasks grow but economics remain customer- and integration-specific; the bear case assumes demonstrations do not translate into productive hours, hardware-price compression outpaces software or service capture, and new capital arrives at a lower mark. The retained comparable signals are directional only: Galbot's disclosed $3bn financing valuation, Unitree's reported approximately $6.2bn IPO target, and the market-scale forecasts from Goldman Sachs, CNBC, TrendForce, and IDTechEx measure different companies, stages, units, and denominators. They cannot be applied as a revenue multiple to Stardust. Exit readiness is consequently low-to-moderate: a future strategic or public-market exit would require audited financials, customer references, delivered rather than ordered units, repeatable gross margin, actuator MTBF, safety records, and a clean cap table showing liquidation preferences and dilution. Until then, track is preferable to underwriting a target return from an unsupported mark.[CV011, CV012, CV013, CV014, CV015]

8.4 Exhibits