DexRobot presented its full range of dexterous robotic hands and introduced the DexTele Teleoperation Data Acquisition System at Automate 2026 in Chicago, where the event placed increased emphasis on humanoid robotics and artificial intelligence applications in industrial automation. The company’s DexHand product line includes systems aimed at research, manufacturing and education markets. According to DexRobot, the flagship DexHand021 Pro uses a dual-tendon design and provides 22 degrees of freedom, a payload capacity of 50 newtons, full-palm multimodal sensing and durability exceeding 300,000 operating cycles. The company stated that the system is priced at approximately one-fifth the cost of comparable products.
DexRobot also highlighted the DexHand021, an industrial-oriented model with 19 degrees of freedom, 23 integrated sensors and a tendon-driven underactuated design intended to support stable grasping and tool-handling tasks. The entry-level DexHand021 S offers eight degrees of freedom, weighs 0.6 kilograms and includes an open software development kit. The company said the model is designed to provide core hand functionality at a lower price point.
The company also introduced the DexTele Teleoperation Data Acquisition System, which combines teleoperation hardware, force-controlled humanoid robotic arms, dexterous hands and a data management platform. DexRobot said the system is designed for deployment without additional development work related to structural components, electrical systems, communications, teleoperation algorithms or data-processing pipelines.
According to the company, the platform supports teleoperation and multimodal data collection across a range of complex movements with low latency and high precision. Supported data types include tactile, joint, pose, vision and depth information.
DexRobot said the DexHand and DexTele products are intended to function as an integrated ecosystem combining hardware, data collection and software tools. The company stated that the offering is designed to support AI-enabled robotic manipulation and the generation of training data for automation applications across manufacturing, research and educational environments.
