RoboROI

Region Germany France United Kingdom United States Every region with sourced labour data is offered. 3 of the 4 cannot yet be computed end to end (Germany, France, United Kingdom); pick one and the page says exactly which figure is missing rather than filling it in.

Industry Manufacturing Retail Warehousing Sets which occupation the robot is compared against — each is a different OEWS wage, not a relabelling of one figure.

Robot model Select a model… 1X Technologies — NEO (no published price) Agility Robotics — Digit (price unverified) Apptronik — Apollo (no published price) Boston Dynamics — Atlas (no published price) EngineAI — T800 (no published price) Figure AI — Figure 02 (no published price) Fourier Intelligence — GR-2 (no published price) Sanctuary AI — Phoenix (no published price) Tesla — Optimus (no published price) UBTECH — Walker S (no published price) Unitree Robotics — G1 Unitree Robotics — H1 Unitree Robotics — H2 Every model in the dataset is listed. No manufacturer is pre-selected.

Number of robots

Shifts per day Up to 3 × 8-hour shifts — a robot cannot work more of a day than exists. Availability within those hours is the uptime assumption, shown with the results.

Productivity vs a human worker Conservative — 60% Moderate — 72% Aggressive — 85% Named presets rather than a slider: these figures have no published source, and a slider would invite false precision on the most debatable assumption in the model.

Horizon Years to model, 1–10. Beyond about five years the hardware assumptions stop meaning much for a product category this young.

Include hidden labour costs Turnover, absenteeism, workers' compensation and HR administration, as one blended uplift on human labour cost. Off by default and unverified — this figure has no source, and switching it on makes the robot's case look better, so the caveat travels with it onto the results page. Free to use; there is no paywall in this tool.