Chapter 12: What Manufacturers Must Own

Overview

Vendors may own general models, but manufacturers must own process truth, evaluation harnesses, failure taxonomies, data rights, and rollback authority. This chapter rewrites the problem as a factory-cell data contract. The promise of large robot datasets in ^[1] and ^[2] matters, but in manufacturing an episode becomes valuable only when it is attached to inspection and process outcomes.

Buy/build decisions should be made by asking who controls episode schema, replay sets, QA traces, safety case, update governance, operator consent, and vendor exit plan, not by demo performance alone. The data is therefore not just a camera stream. It includes hand choice, force-torque, tactile patches, controller mode, operator intervention, and inspection result. ^[3] and ^[4] show why human-to-robot transfer loses critical signals unless the capture interface is designed around embodiment and contact.

After reading this chapter... - Explain factory manipulation as a data contract rather than a model score. - Separate two-finger, suction, custom, and five-finger choices by collection cost and failure observability. - Read papers, company releases, and deployment claims through evidence tiers. - Design the replay set and QA trace required for a first manufacturing PoC.

Core Map

Visual Argument

Figure 12.1. Factory loop as ownership map. Source: reused local survey asset or author-created illustration.

Figure 12.2. Readiness matrix for choosing cells and vendors. Source: reused local survey asset or author-created illustration.

Figure 12.3. Operations layer after model deployment. Source: reused local survey asset or author-created illustration.

Six Assets Manufacturers Must Own

The practical meaning of six assets manufacturers must own is that process variables must be left in a learnable form. ^[1] shows how scale can broaden embodiment and task families, yet a factory cell is narrower and stricter. A pick that looks identical to a person can become a different distribution when fixture tolerance, surface contamination, cycle pressure, or reject code changes.

In the concrete scenario, Buy/build decisions should be made by asking who controls episode schema, replay sets, QA traces, safety case, update governance, operator consent, and vendor exit plan, not by demo performance alone. Human demonstrations alone are not enough. Even when an interface such as ^[2] makes collection easier, a policy will repeat the same release failure if contact force and failure cause are absent. The episode schema must bind observation, action, contact state, QA outcome, and operator note under one key.

The evidence tier matters. A source such as ^[3] gives a method and benchmark that can be inspected. A company source can reveal deployment direction, but usually exposes less about data rights, recovery handling, and operating metrics. Manufacturers should place both in the same comparison table but never read them with the same confidence.

From a control perspective, the learned policy should not be the whole system. Force limits, guarded motion, fixture state, collision zones, and rollback conditions have to sit around the model. Without that boundary, collecting more data can create more safety stops and more rework instead of a more deployable robot.

From an operating perspective, the reproducibility of failure matters more than a headline success rate. A failure has to enter a replay set, the next update has to pass that replay set, and the deployed cell has to show a lower rate for the same failure code. The question is not only how to describe six assets manufacturers must own, but which missing log would make the claim collapse on a line.

The Boundary Between Vendor Models and Process Truth

The practical meaning of the boundary between vendor models and process truth is that process variables must be left in a learnable form. ^[4] shows how scale can broaden embodiment and task families, yet a factory cell is narrower and stricter. A pick that looks identical to a person can become a different distribution when fixture tolerance, surface contamination, cycle pressure, or reject code changes.

In the concrete scenario, Buy/build decisions should be made by asking who controls episode schema, replay sets, QA traces, safety case, update governance, operator consent, and vendor exit plan, not by demo performance alone. Human demonstrations alone are not enough. Even when an interface such as ^[5] makes collection easier, a policy will repeat the same release failure if contact force and failure cause are absent. The episode schema must bind observation, action, contact state, QA outcome, and operator note under one key.

The evidence tier matters. A source such as ^[6] gives a method and benchmark that can be inspected. A company source can reveal deployment direction, but usually exposes less about data rights, recovery handling, and operating metrics. Manufacturers should place both in the same comparison table but never read them with the same confidence.

From a control perspective, the learned policy should not be the whole system. Force limits, guarded motion, fixture state, collision zones, and rollback conditions have to sit around the model. Without that boundary, collecting more data can create more safety stops and more rework instead of a more deployable robot.

From an operating perspective, the reproducibility of failure matters more than a headline success rate. A failure has to enter a replay set, the next update has to pass that replay set, and the deployed cell has to show a lower rate for the same failure code. The question is not only how to describe the boundary between vendor models and process truth, but which missing log would make the claim collapse on a line.

Evaluation Harnesses and Replay Sets

The practical meaning of evaluation harnesses and replay sets is that process variables must be left in a learnable form. ^[7] shows how scale can broaden embodiment and task families, yet a factory cell is narrower and stricter. A pick that looks identical to a person can become a different distribution when fixture tolerance, surface contamination, cycle pressure, or reject code changes.

In the concrete scenario, Buy/build decisions should be made by asking who controls episode schema, replay sets, QA traces, safety case, update governance, operator consent, and vendor exit plan, not by demo performance alone. Human demonstrations alone are not enough. Even when an interface such as ^[8] makes collection easier, a policy will repeat the same release failure if contact force and failure cause are absent. The episode schema must bind observation, action, contact state, QA outcome, and operator note under one key.

The evidence tier matters. A source such as ^[9] gives a method and benchmark that can be inspected. A company source can reveal deployment direction, but usually exposes less about data rights, recovery handling, and operating metrics. Manufacturers should place both in the same comparison table but never read them with the same confidence.

From a control perspective, the learned policy should not be the whole system. Force limits, guarded motion, fixture state, collision zones, and rollback conditions have to sit around the model. Without that boundary, collecting more data can create more safety stops and more rework instead of a more deployable robot.

From an operating perspective, the reproducibility of failure matters more than a headline success rate. A failure has to enter a replay set, the next update has to pass that replay set, and the deployed cell has to show a lower rate for the same failure code. The question is not only how to describe evaluation harnesses and replay sets, but which missing log would make the claim collapse on a line.

Data Rights and Worker Consent

The practical meaning of data rights and worker consent is that process variables must be left in a learnable form. ^[10] shows how scale can broaden embodiment and task families, yet a factory cell is narrower and stricter. A pick that looks identical to a person can become a different distribution when fixture tolerance, surface contamination, cycle pressure, or reject code changes.

In the concrete scenario, Buy/build decisions should be made by asking who controls episode schema, replay sets, QA traces, safety case, update governance, operator consent, and vendor exit plan, not by demo performance alone. Human demonstrations alone are not enough. Even when an interface such as ^[11] makes collection easier, a policy will repeat the same release failure if contact force and failure cause are absent. The episode schema must bind observation, action, contact state, QA outcome, and operator note under one key.

The evidence tier matters. A source such as ^[12] gives a method and benchmark that can be inspected. A company source can reveal deployment direction, but usually exposes less about data rights, recovery handling, and operating metrics. Manufacturers should place both in the same comparison table but never read them with the same confidence.

From a control perspective, the learned policy should not be the whole system. Force limits, guarded motion, fixture state, collision zones, and rollback conditions have to sit around the model. Without that boundary, collecting more data can create more safety stops and more rework instead of a more deployable robot.

From an operating perspective, the reproducibility of failure matters more than a headline success rate. A failure has to enter a replay set, the next update has to pass that replay set, and the deployed cell has to show a lower rate for the same failure code. The question is not only how to describe data rights and worker consent, but which missing log would make the claim collapse on a line.

A 2026-2030 Execution Roadmap

The practical meaning of a 2026-2030 execution roadmap is that process variables must be left in a learnable form. ^[13] shows how scale can broaden embodiment and task families, yet a factory cell is narrower and stricter. A pick that looks identical to a person can become a different distribution when fixture tolerance, surface contamination, cycle pressure, or reject code changes.

In the concrete scenario, Buy/build decisions should be made by asking who controls episode schema, replay sets, QA traces, safety case, update governance, operator consent, and vendor exit plan, not by demo performance alone. Human demonstrations alone are not enough. Even when an interface such as ^[15] makes collection easier, a policy will repeat the same release failure if contact force and failure cause are absent. The episode schema must bind observation, action, contact state, QA outcome, and operator note under one key.

The evidence tier matters. A source such as ^[15] gives a method and benchmark that can be inspected. A company source can reveal deployment direction, but usually exposes less about data rights, recovery handling, and operating metrics. Manufacturers should place both in the same comparison table but never read them with the same confidence.

From a control perspective, the learned policy should not be the whole system. Force limits, guarded motion, fixture state, collision zones, and rollback conditions have to sit around the model. Without that boundary, collecting more data can create more safety stops and more rework instead of a more deployable robot.

From an operating perspective, the reproducibility of failure matters more than a headline success rate. A failure has to enter a replay set, the next update has to pass that replay set, and the deployed cell has to show a lower rate for the same failure code. The question is not only how to describe a 2026-2030 execution roadmap, but which missing log would make the claim collapse on a line.

Manufacturing Cell Checkpoint

Start a PoC by writing the data contract before selecting the model. Buy/build decisions should be made by asking who controls episode schema, replay sets, QA traces, safety case, update governance, operator consent, and vendor exit plan, not by demo performance alone. The contract should include episode ID, operator or teleop ID, hand or tool ID, contact channel, controller mode, quality decision, defect code, override reason, and rollback condition. If a vendor hides or cannot export these fields, the manufacturer cannot explain why model improvement happens.

The second checkpoint is evidence tiering. Papers with arXiv or DOI links support method claims; official company pages support product-direction claims; press and watchlist items should not become load-bearing claims. PI, Generalist, Skild, Figure, Covariant, Dexterity, Chef Robotics, Sunday, Config, and CarbonSix all fit under data-driven manipulation, but they expose very different evidence about manufacturing readiness.

Open Questions and Failure Modes

First, large data without contact state leaves insertion, wiping, deformable handling, and tool use under-observed. Second, when data rights stay entirely with the vendor, the manufacturer does not own the cause of process improvement. Third, worker video and process IP are learning assets and governance risks at the same time. Fourth, sim-to-real data does not guarantee factory performance unless it is tied to real QA labels.

Data Contract Addendum

The important unit is not the model name but the observable attempt. To a human, two factory cycles can look like the same task; to a robot, they may differ in surface friction, initial pose, contact order, force limit, and inspection rule. Data strategy is therefore an operating system for connecting failures to process variables, not a campaign for collecting success clips.

Factory data is slower and messier than benchmark data, but it carries more decisive labels. Defect codes, rework, operator intervention, line stops, and safety stops are not noise to be hidden from learning. They are the labels that decide whether a policy can be deployed. If those labels are detached from episodes, model curves can improve while process KPIs stay flat.

Large-data driven manipulation is therefore not a choice of one VLA. It is a joint design of task schema, hand choice, controller boundary, replay set, QA trace, and update governance. Without that design, a foundation model may produce impressive demonstrations without becoming repeatable replacement labor in manufacturing.

Operating Loop Addendum

The important unit is not the model name but the observable attempt. To a human, two factory cycles can look like the same task; to a robot, they may differ in surface friction, initial pose, contact order, force limit, and inspection rule. Data strategy is therefore an operating system for connecting failures to process variables, not a campaign for collecting success clips.

Factory data is slower and messier than benchmark data, but it carries more decisive labels. Defect codes, rework, operator intervention, line stops, and safety stops are not noise to be hidden from learning. They are the labels that decide whether a policy can be deployed. If those labels are detached from episodes, model curves can improve while process KPIs stay flat.

Large-data driven manipulation is therefore not a choice of one VLA. It is a joint design of task schema, hand choice, controller boundary, replay set, QA trace, and update governance. Without that design, a foundation model may produce impressive demonstrations without becoming repeatable replacement labor in manufacturing.

Deployment Governance Addendum

The important unit is not the model name but the observable attempt. To a human, two factory cycles can look like the same task; to a robot, they may differ in surface friction, initial pose, contact order, force limit, and inspection rule. Data strategy is therefore an operating system for connecting failures to process variables, not a campaign for collecting success clips.

Factory data is slower and messier than benchmark data, but it carries more decisive labels. Defect codes, rework, operator intervention, line stops, and safety stops are not noise to be hidden from learning. They are the labels that decide whether a policy can be deployed. If those labels are detached from episodes, model curves can improve while process KPIs stay flat.

Large-data driven manipulation is therefore not a choice of one VLA. It is a joint design of task schema, hand choice, controller boundary, replay set, QA trace, and update governance. Without that design, a foundation model may produce impressive demonstrations without becoming repeatable replacement labor in manufacturing.

Field Failure Addendum

The important unit is not the model name but the observable attempt. To a human, two factory cycles can look like the same task; to a robot, they may differ in surface friction, initial pose, contact order, force limit, and inspection rule. Data strategy is therefore an operating system for connecting failures to process variables, not a campaign for collecting success clips.

Factory data is slower and messier than benchmark data, but it carries more decisive labels. Defect codes, rework, operator intervention, line stops, and safety stops are not noise to be hidden from learning. They are the labels that decide whether a policy can be deployed. If those labels are detached from episodes, model curves can improve while process KPIs stay flat.

Large-data driven manipulation is therefore not a choice of one VLA. It is a joint design of task schema, hand choice, controller boundary, replay set, QA trace, and update governance. Without that design, a foundation model may produce impressive demonstrations without becoming repeatable replacement labor in manufacturing.

What to Learn Next

The next chapter applies the same principle to hands and end-effectors. Some cells are best served by suction or two fingers; others cannot close the data loop without five-finger hands and tactile or force-rich data.

References

1. O'Neill, Abby (2023). Open X-Embodiment: Robotic Learning Datasets and RT-X Models. arXiv.
2. Khazatsky, Alexander (2024). DROID: A Large-Scale In-The-Wild Robot Manipulation Dataset. arXiv.
3. AgiBot World Team (2025). AgiBot World Colosseo: A Large-Scale Manipulation Platform. arXiv.
4. Ha, Huy (2024). Universal Manipulation Interface: In-The-Wild Robot Teaching Without In-The-Wild Robots. arXiv.
5. Choi, Hojung (2026). In-the-Wild Compliant Manipulation with UMI-FT. arXiv.
6. Black, Kevin (2024). pi0: A Vision-Language-Action Flow Model for General Robot Control. arXiv.
7. Physical Intelligence (2025). OpenPI: Open Source Robot Policy Stack. GitHub.
8. Covariant (2024). RFM-1: Robotics Foundation Model. Company technical post.
9. Dexterity (2025). Dexterity Foresight: AI Platform for Industrial Robot Workcells. Company product page.
10. Chef Robotics (2025). ChefOS: AI Robotics Platform for Food Manufacturing. Company product page.
11. Figure AI (2026). Figure 03 + Helix 02: General-Purpose Humanoid System. Company product page.
12. Generalist AI (2025). GEN-0 Robot Foundation Model. Company page.
13. Skild AI (2024). General-Purpose Robot Brain. Company page.
14. NVIDIA (2025). Isaac Lab: A GPU-Accelerated Simulation Framework for Multi-Modal Robot Learning. NVIDIA Research.
15. NVIDIA (2025). Isaac GR00T N1 Open Foundation Model for Humanoid Robots. NVIDIA Developer.
16. Toyota Research Institute (2024). Large Behavior Models for Robot Manipulation. Company technical post.
17. Li, Xingyu (2024). Evaluating Real-World Robot Manipulation Policies in Simulation. arXiv.
18. Mandlekar, Ajay (2021). What Matters in Learning from Offline Human Demonstrations for Robot Manipulation. arXiv.
19. Zhao, Tony Z. (2023). Learning Fine-Grained Bimanual Manipulation with Low-Cost Hardware. arXiv.
20. Chi, Cheng (2023). Diffusion Policy: Visuomotor Policy Learning via Action Diffusion. arXiv.
21. Yu, Wenhao (2025). ForceVLA: Enhancing VLA Models with a Force-aware MoE for Contact-rich Manipulation. arXiv.
22. Huang, Yuhang (2025). Tactile-VLA: Unlocking Vision-Language-Action Model's Physical Knowledge for Tactile Generalization. arXiv.
23. Shaw, Kenneth (2023). LEAP Hand: Low-Cost, Efficient, and Anthropomorphic Hand for Robot Learning. arXiv.
24. Lambeta, Mike (2024). Digitizing Touch with an Artificial Multimodal Fingertip. arXiv.