Geekplus, the mobile robotics specialist, has been named a winner of the 2026 RBR50 Innovation Award for its Robot Arm Picking Station, following a successful deployment at Schneider Electric’s Shanghai warehouse
The RBR50 awards, presented annually by The Robot Report at the Robotics Summit & Expo, recognise the 50 most innovative companies in global robotics. This marks the fifth time Geekplus has received the accolade, placing it alongside the likes of ABB, Amazon, Boston Dynamics and Nvidia.
The award was given specifically for the Robot Arm Picking Station’s ability to automate what remains one of the most persistent manual bottlenecks in warehouse operations: item picking. While autonomous mobile robots (AMRs) have transformed storage and transport within distribution centres, the picking process itself has largely resisted automation. Rapidly changing SKU catalogues, diverse product forms and the high training costs associated with conventional vision models have made it one of the last major hurdles standing in the way of fully automated fulfilment.
The RBR50 judging panel praised the solution for closing this gap within Geekplus’ existing goods-to-person portfolio, noting that by embedding robotic picking into its ecosystem, Geekplus moves closer to fully autonomous, end-to-end warehouse operations.
The Robot Arm Picking Station tackles the challenge by combining Geekplus’ proprietary embodied intelligence foundation model, Geekplus Brain, with zero-shot learning technology. This enables robotic arms to pick accurately across large-scale SKU ranges without requiring per-item training, a significant departure from conventional robotic picking systems that demand extensive programming for each new product type. The station integrates directly with existing AMR infrastructure, enabling a complete unmanned picking workflow from storage through to dispatch.
In the Schneider Electric pilot, the results were compelling. The system doubled manual throughput rates, achieved picking accuracy of 99.99% or higher, reaching 100% during testing, and was production-ready within just 48 hours of deployment. Because the station is pre-trained on large-scale real-world data via Geekplus Brain, it requires no secondary training on site and can dynamically adapt to business fluctuations and packaging changes without intervention. All data processing remains on-premise, addressing a key concern for manufacturers and logistics operators handling sensitive operational information.
Geekplus’ unified All-in-One software architecture also allowed seamless integration with Schneider Electric’s existing warehouse network and production processes, without requiring changes to the facility’s established workflows. This plug-and-play capability is expected to make the solution attractive to warehouse operators looking to introduce robotic picking without overhauling their current systems.
The company says the successful deployment establishes a replicable model for embodied intelligence in industrial warehousing and represents a meaningful step toward fully unmanned warehouse operations. Geekplus plans to expand the solution’s application across additional industries and use cases as it continues to build out its end-to-end automation offering.
