The cell supports development of manipulation and inspection systems: compliant arms, force-controlled end-effectors, vision rigs and the mock-ups that let a system fail somewhere that does not matter.
That private failure phase is the facility's real offering. A robot learning a real task will drop things, miss things and behave unexpectedly, and doing that on a production line destroys the goodwill a deployment depends on. Doing it here, against a representative mock-up, costs nothing but time.
Safety assessment is treated as part of the engineering. The facility's practice is to involve a safety specialist early, on the basis that a cell designed without one either gets redesigned later or stays permanently behind a fence.
The cell also supports comparative studies — a robot and a human inspector performing the same task on the same specimens — which is the evidence regulated industries actually need and which is surprisingly hard to arrange elsewhere.
Specification
- Collaborative and industrial manipulators with force control
- Compliant and custom end-effector development capability
- Machine vision rigs with controlled and adversarial lighting
- Mock-up construction area for representative task environments
- Motion capture for ground-truth measurement of manipulation
- Instrumented test specimens for repeatable inspection trials
What it is the right answer to
Jobs this facility genuinely suits.
- Developing a manipulation task before it goes near production
- Comparative inspection studies against a human baseline
- End-effector design and force-limit characterisation
- Perception development under realistic and deliberately poor lighting
- Safety assessment and risk reduction for a planned industrial cell
What to bring
Turning up without these costs facility hours rather than bench hours.
- The task, demonstrated by several different people so the variation is visible
- Representative parts or specimens, including the awkward ones
- A defect set that is not a stock photograph, where inspection is the goal
- Willingness to let the system fail repeatedly before anyone draws conclusions
What you leave with
- A working development system with characterised performance and failure modes
- Comparative data against the human baseline, reported honestly
- Safety assessment documentation for the intended deployment
- A clear statement of what remains unproven outside the cell
Limitations
Worth knowing before you write the proposal.
- A mock-up is representative, not real; site commissioning remains necessary
- Cycle times achieved in development are optimistic relative to production
- Comparative studies need enough specimens to be statistically meaningful