Research and pilot
reRobotics
reRobotics develops AI-guided, non-destructive disassembly and refurbishment for retired robots. We are building it with university and research partners, and plan a US base to take retired fleets apart, recalibrate them and return them to work.
Why it matters
Warehouse, factory and service robots are retiring in growing numbers. Their servo motors, reducers, lidar, controllers and battery packs keep most of their value, but only if they come out intact and go back in calibrated.
The repair, refurbishment and recycling market for logistics robots was an estimated $1.9–2.6 billion in 2023 and is projected to reach $8–11 billion by 2030. Refurbished units typically sell at 40–60% of the price of new.
Generic scrappers strip robots by weight. Precision gearboxes, connectors and sensors are damaged on the way out, and OEM software locks turn good parts into scrap.
Retired robots carry site maps, edge models and control firmware. US buyers increasingly expect them to be wiped, re-certified and traced inside the country.
Market figures are Circunar estimates (2025).
The system
Each refurbishment cell combines multimodal perception, robotic disassembly tools and a digital passport for every unit, coordinated across cloud, local and edge layers.
Product digital passport with disassembly difficulty, condition and bill of materials; AI model management with online and offline learning; task scheduling, monitoring and access control.
Local data preprocessing, safety management, task allocation and data relay. Simulation and augmented reality validate each procedure before it runs.
Path planning and task coordination for the disassembly arms, with state management, position monitoring and encrypted telemetry.
Robot arms and disassembly tools, RGB-D cameras, LIBS spectral analyzers for material identification, safety light curtains and thermal and smoke sensors.
Screws, connectors and wire harnesses are classified and their 6D pose estimated from RGB, depth and point-cloud data, so the arm can reach and release them without damage.
Models train on rendered synthetic datasets and real captures from the cell, then fine-tune as each new robot model arrives.
LIBS spectral analysis identifies alloys and materials, so every recovered part goes to the right reuse or recycling stream.
Refurbishment scope
Rewinding, bearing replacement, backlash and torque calibration
Recalibration and re-certification
Firmware reflash and full data wipe
Grading and BMS rebalancing, routed through reBattery
Inspection, wear-part replacement and refinishing
Refurbished robots go to buyers who could not afford new ones, such as smaller manufacturers, farms and regional warehouses, or back into Circunar-managed fleets under CPO standards.
Research partnership
Circunar works with CUNY, RPI and other research institutions on the hard problems of robotic disassembly: perception, manipulation planning, remaining-useful-life estimation for motors and gearboxes, and training for refurbishment technicians.
Joint research projects and a pilot disassembly cell with university partners.
A US disassembly and refurbishment base (planned), starting with retired warehouse drive units.
Extend to multi-axis arms and service robots, and feed certified units into CPO and second-life markets.
Research institutions, fleet owners and OEMs: tell us about your robots, lab or retired inventory.