B02
Dual-arm picking cell
A mobile base with a lift and two arms navigates aisles, reads SKUs from labels, and picks boxes and irregular items from shelves onto a sortation cart.
From digital twins and industrial cells to robot learning and multi-robot control, across the full lifecycle: concept, design, simulation, perception, deployment.
Each one modelled, simulated and validated end to end before anything moves on a real floor.
A mobile base with a lift and two arms navigates aisles, reads SKUs from labels, and picks boxes and irregular items from shelves onto a sortation cart.
A full material-handling cell: a box rides the conveyor, an arm places it on a pallet, a forklift carries it to the shelf — assembled from a scene manifest, run with one command.
A conformance-tested arm and gripper simulation built against a published hardware interface, with the ROS 2 nodes to drive it and a mock layer for contract testing.
A tool-driven control layer that lets an operator direct cameras, lighting, transforms and keyframes inside a running production scene from a single interface.
Concept and mechanical design through simulation, perception and deployment — so the handoffs that usually lose weeks don't exist.
Factories, facilities and equipment rebuilt as physically accurate environments — dual-fidelity assets that render and simulate from one source of truth.
Arms, conveyors and full lines: motion planning, grasping and the behaviour logic that holds a cell together under real cycle times.
Perception pipelines, synthetic data generation, and policy training and evaluation — reinforcement, imitation and vision-language-action models.
Navigation stacks, fleet coordination and cloud control, containerised so the same build runs in simulation and on the robot.
We work with startups, enterprises and research teams across manufacturing, logistics and healthcare — from a single simulation study to a full build.
info@kemabots.com