Five-Finger Gripper Joins TM Arm for Flexible Production

High-mix production just got a five-finger grip on reality.

Techman Robot Korea and Tesollo unveiled a joint system at AW 2026 that pairs a humanoid hand with a TM cobot arm, pitched for high-mix, low-volume manufacturing. The demo spotlighted a Tesollo end effector mounted on a TM arm, with the parties highlighting the appeal of combining industrial reliability with adaptive grasping to tackle tasks conventional automation struggles with. The companies have a history of collaboration: they previously demonstrated teleoperation tying a humanoid hand to a TM arm, and Tesollo has since commercialized its DG-5F-S five-fingered gripper. The latest showpiece appears to be a pathway from teleoperation toward more autonomous, dexterous handling in a production environment.

The open technical specifics leave some gaps, but the picture is clear enough to draw the line from concept to pragmatism. The demonstration setup referenced a TM cobot arm working with Tesollo hardware—initially described as a three-finger gripper in the AW 2026 footage. Tesollo’s DG-5F-S five-finger gripper was highlighted in accompanying notes as a more mature, commercially available option. In practice, a Delto gripper variant is mentioned as an interface on the TM arm, with additional vision capability contemplated: “a camera can be mounted underneath the Delto gripper on the TM arm.” The fingertips, whatever their exact configuration, are part of an architecture designed to enable flexible grasping across a range of parts.

DoF counts for the humanoid elements are not disclosed in the available materials. The DG-5F-S is described as a five-fingered gripper, but the article does not publish a formal degrees-of-freedom tally for the hand, nor does it specify the TM arm’s joint configuration in this specific integration. For readers and buyers, that omission matters: dexterous grasping at scale hinges on finger joint count, tendon routing, and control bandwidth. Industry practice would expect multi-DOF fingers (often 15–20 joints total across a five-finger hand in similar demonstrations), but until Tesollo or Techman publishes precise figures, the DOF remains an uncertainty. The same caution applies to the end-to-end kinematic range and repeatability specifications.

Payload and power notes are more concrete. The system is reported to support a maximum payload of 20 kg (44 lb) with the TM arm-Delto configuration. That payload ceiling places the platform in the realm of light-to-medium components, suitable for bin-picking, assembly, and packaging tasks common in consumer electronics, automotive trims, and logistics—but not for heavy machining fixtures or large mechanical parts. Power, runtime, and charging requirements were not disclosed in the coverage, leaving a critical gap for ROI and lifecycle planning.

Technology readiness is best described as a lab/demo proposition, with the public display at a trade show acting as a proof-of-concept for a configurable end-effector on a TM arm. The message is less about field deployment and more about demonstrating a pathway toward flexible grasping in environments where part variety and small lot sizes complicate traditional automation. In other words, a credible step forward, but not a ready-to-run production line solution—yet.

Two practitioner-level takeaways stand out. First, for high-mix, low-volume production, the ability to swap a dexterous end effector on a familiar robot arm can dramatically cut changeover times and footprint compared with fixed-geometry grippers. The DG-5F-S five-finger design hints at broader adaptability across part geometries, but actual reliability across hundreds of SKUs remains to be demonstrated in the field. Second, the teleoperation lineage matters: moving from controlled-teleop demonstrations to autonomous or semi-autonomous grabbing requires robust control loops, tactile feedback, and integrated sensing. The lack of published endurance, precision, and safety metrics means engineers should watch for repeatability tests, cycle-time data, and fault-handling software in future disclosures.

Compared to the prior generation—where the emphasis was on teleoperation bridging humanoid hands and TM arms—the current path leans toward deploying a more capable, fixed-end-effector on a standard industrial arm, with vision and sensing options that could enable semi-autonomy. The improvement is meaningful, but the story isn’t finished: till full specs, reliability metrics, and field-test results surface, claims of “industrial reliability meets adaptive grasping” remain aspirational rather than prescriptive.

What to watch next: the exact DOF specification for the end effector, published endurance and cycle-time benchmarks, and a clear power/charging profile. If Tesollo and Techman publish concrete figures—especially DOF counts, repeatability (in mm), and sustained runtime under typical bin-picking workloads—their system could shift from a compelling concept to a practical, flexible automation backbone for small-batch manufacturers.

Sources

Tesollo and Techman Robot unveil robot for high-mix, low-volume production