ReconCell: A Reconfigurable Robot Workcell

The reconﬁgurable robot workcell in a conﬁguration where it assembles an automotive headlight.

ReconCell: A Reconfigurable Robot Workcell

ReconCell was a reconfigurable robot workcell conceived to give low-volume, high-mix manufacturers the agility of large-scale automation without the overhead. The project unites several complementary research threads: smart, passive hardware modules that can be rearranged in minutes; a ROS-based, robot-agnostic control stack that couples programming-by-demonstration with [meta-scripted state-machine generation (SMACHA) for rapid task authoring; kinesthetic-guided base-frame calibration that lets multiple collaborative robots share a common frame with millimetre accuracy; and intuitive tools for quality control, part localisation and process monitoring. Tested across automotive, electronics and bespoke assembly lines, ReconCell shows that flexible automation can be precise, economical and user-friendly—setting the stage for SME-friendly commercial deployment of truly adaptable robot workcells.

Roles

Nov 1 2015 - Feb 28 2019: Senior Assistant | Postdoc. @ JSI

Designed SMACHA API for ROS state machine scripting & code generation.
Co-designed ROS architecture for control of rapidly reconfigurable robotic workcell.
Coordinator between robot control team & visual programming interface team.