Proxie Gen 2 brings two armed mobile manipulation

Twenty-eight Proxie robots have logged nearly 13,000 field hours.

Collaborative Robotics, known as Cobot, unveiled Proxie Gen 2 at Automate 2026, pitching it as a production-proven mobile manipulator that finally packs mobility and dexterity into one end-to-end platform. The upgrades include autotasking, autonomous task identification, self-swapping batteries, and a new two-armed manipulation option, along with a stated increase in payload capacity. The company says the Gen 2 design is aimed at expanding deployments across healthcare, logistics, and manufacturing, where the combination of moving around a facility and handling objects with precision is repeatedly the bottleneck.

Testing shows that field experience is where Proxie’s improvements prove themselves. The Robot Report notes that 28 Proxie robots have accumulated nearly 13,000 operating hours, traveled more than 22,000 miles, moved over 154,000 carts, and saved workers millions of steps. Those numbers come from real-world deployments rather than lab bench tests, a shift Cobot emphasizes as essential before marketing the platform as production-ready. Porter’s emphasis on field readiness underlines a broader industry discipline: you don’t know where the stresses will crack hardware until hardware meets real workflows over months of use. The Mayo Clinic has been one of the notable deployments, with Proxie used to transport materials around clinical spaces, a use case that highlights both the promise and the ongoing safety considerations in patient-adjacent environments.

Two key capabilities define Gen 2’s edge, according to the company. Autotasking and autonomous task identification aim to cut software integration friction, letting the robot decide which tasks to pursue within a defined workflow rather than requiring bespoke control scripts for each job. Self-swapping batteries address uptime, a perennial constraint for mobile manipulators operating across expansive facilities, hospitals, or labs where downtime can cascade into missed deliveries or disrupted lines. The new two-armed option expands dexterity without requiring separate robots or bespoke end effectors for each task, although it also increases the system’s mechanical complexity and maintenance footprint. The company reports that Gen 2’s design is intended to be production-proven, a claim borne out by the breadth of live environments in which the platform has already operated.

From an industrial perspective, Gen 2 represents a deliberate move to reduce the total cost of ownership for mobile manipulation. The combination of autonomous task selection and a dual-armed config helps address two stubborn realities: the pace of deployment in busy facilities and the diverse handling duties across departments. Yet observers note that real-world ROI will hinge on maintaining reliability of perception and pick-and-place in cluttered spaces, ensuring battery swaps align with task cadence, and managing the additional calibration and maintenance that come with a two-armed payload. In healthcare settings especially, safety, cleaning regimes, and tool compatibility will be constant tests for any material-handling robot.

Looking ahead, industry watchers will want to see more production-level metrics from additional sites, including longer-term uptime and the cost profile of adding two arms versus upgrading single-arm variants with better grippers. Partners will be watching whether autotasking reduces the need for custom software connectors and whether autonomous task identification holds up as workflows scale from pilots to multi-site productions. If Gen 2 can consistently deliver on field performance while maintaining low friction for facilities teams, it could become a steadier, more widely adopted platform for mobile manipulation across hospitals, warehouses, and factories.