Explore a supply chain
Start with a component in the manufacturing explorer. Filter by product and country, then open a supplier record for its listed location, products, and customer context.
Treat headquarters, production location, customer location, and listing country as different facts. A shared product category establishes a comparable activity, not a verified supplier relationship. Named links in a supply chain need direct evidence.
| Dependency to investigate | Research path | Verification question |
|---|---|---|
| Motion components | Muscle → Suppliers | What specifications, qualification, and production capacity are disclosed? |
| Sensors and feedback | Senses → Suppliers | Is the part qualified for the named customer’s task? |
| Power and runtime | Heart → Suppliers | Under what duty cycle were runtime and thermal limits measured? |
| Deployment and support | Integration → Companies | Who installs, maintains, and recovers the system? |
Explore how a business earns revenue
These are questions to guide reading, not automatic labels assigned to companies.
| Revenue model | Research path | What to examine |
|---|---|---|
| Components | Muscle, Senses, Heart | Unit economics, qualification, replacement cycles, customer concentration |
| Complete systems | Integration | Delivery acceptance, installation cost, warranty and support obligations |
| Models and software | Brain, Nerves | Paid licensing, distribution, switching costs, competition from bundled software |
| Integration and maintenance | Integration | Service terms, labor requirements, repeatability across sites |
| Consumables and task-based services | Work | Usage, recurring purchases, customer outcomes, support cost |
| Data and evaluation | Blood, Immune | Data rights, independent testing, buyer trust, willingness to pay |
Compare two possible industry structures
The following scenarios are research questions. Neither is assigned a probability or used to change index weights.
| Reusable models across robot bodies | Closely integrated hardware and software | |
|---|---|---|
| Evidence that would support it | Reliable transfer to different robots with limited adaptation; independently paid model licenses | Repeated customer value tied to a specific body, controller, and proprietary deployment data |
| What to measure | Adaptation cost, task performance, support burden, license economics | Full-system economics, fleet utilization, retention, improvement across deployments |
| What would weaken it | Repeated customization and support consume the license margin | Comparable alternatives match performance with lower integration cost |
Both structures may coexist in different tasks. Use commercial evidence to test them rather than treating an analogy as a forecast.
Read cost estimates carefully
A bill of materials describes a particular machine at a particular date and production assumption. Check whether software development, integration, maintenance, and human assistance are excluded. We do not treat one robot’s cost estimate as an industry-wide cost breakdown.