Repetitive, dexterous handling - kitting and sorting from messy bins, manipulating deformable parts like seals and gaskets, routing cables and hoses. If a person can show it in a few minutes, Ludis can capture it.

It depends on the task's variability, but most pilots collect demonstrations over a few weeks at 1–2 hours a day. More demonstrations of the edge cases make the policy more robust.

Yes. The gripper that records is the gripper that runs - mount it on standard cobot arms like Universal Robots or xArm, or on AGVs for mobility. Your data isn't tied to our hardware.

No. Operators record during normal shifts with handheld grippers - there's no teleoperation rig and no robot on the line during capture, so the line keeps moving.

You do. The dataset and the resulting policy are yours to export, retain, and deploy. On-prem and air-gapped capture are available for sensitive sites.

A typical pilot goes from first demonstration to a policy deployed on your arm in roughly 60 days.

Ludis suits repetitive, dexterous handling with real-world variation - kitting and sorting from messy bins, manipulating deformable parts, routing cables and hoses. It is a weaker fit for sub-millimeter rigid assembly or work that fixed automation already does well. The honest way to know is a short pilot on your actual task.

New or changed tasks mean more demonstrations, not re-programming. You record the new variation and the policy learns it - incremental, not a rebuild.

No. Operators collect demonstrations during normal shifts; Ludis processes the sessions and trains the policy. You keep the dataset and the trained model.

Demonstration-trained policies reach roughly 80–100% success in controlled conditions - strong, but not the native six-sigma reliability of hard-coded automation, so human supervision and e-stops stay in place. It doesn't need 99.9% to pay off: a manual line runs at ~80–85% efficiency, while a cobot holding 90–95% across two or three continuous shifts outproduces the manual baseline outright.

The hardware is deliberately open - that's not where the value is. The defensible part is proprietary data scale (large, task-specific demonstration datasets that compound with every site) and deep software integration, including wiring trained policies into your MES/CMMS for predictive maintenance. Anyone can print a gripper; the data and the integration are the moat.

Standard cobot arms - Universal Robots, xArm, Franka and similar - at cobot-safe speeds (~250–1,000 mm/s under ISO/TS 15066) and payloads around 3–35 kg, best on higher-mix work with cycle times over ~10 seconds. Because actions are encoded relative to the gripper and latency-matched at inference, one policy deploys zero-shot across different arms. AGVs add mobility; humanoids are on the roadmap.

A typical cobot cell pays back in about 6–18 months, displacing ~0.5–1.5 full-time-equivalents per shift, lifting overall equipment effectiveness 10–25% and cutting scrap 50–92%. The flip side is why supervision matters: unplanned downtime runs a median near $125,000/hour, so we de-risk with a narrow pilot before scaling.