ORCA: An Open-Source, Reliable, Cost-Effective, Anthropomorphic Robotic Hand for Uninterrupted Dexterous Task Learning

TL;DR

ORCA delivers an open-source, tendon-driven, anthropomorphic hand with 17 DoF, poppable joints, auto-calibration, and integrated tactile sensing at under 2,000 CHF, assembled in hours. It demonstrates reliability and dexterity through long-duration teleoperation, imitation learning, and sim-to-real RL, including multi-hour autonomous runs and tens of thousands of cycles without hardware failure. Key innovations—poppable joints, auto-calibration, and integrated sensing—enable rapid deployment for machine learning-driven dexterous manipulation. By prioritizing cost, maintainability, and accessibility, ORCA aims to broaden participation in dexterous robotic research and real-world applications.

Abstract

General-purpose robots should possess human-like dexterity and agility to perform tasks with the same versatility as us. A human-like form factor further enables the use of vast datasets of human-hand interactions. However, the primary bottleneck in dexterous manipulation lies not only in software but arguably even more in hardware. Robotic hands that approach human capabilities are often prohibitively expensive, bulky, or require enterprise-level maintenance, limiting their accessibility for broader research and practical applications. What if the research community could get started with reliable dexterous hands within a day? We present the open-source ORCA hand, a reliable and anthropomorphic 17-DoF tendon-driven robotic hand with integrated tactile sensors, fully assembled in less than eight hours and built for a material cost below 2,000 CHF. We showcase ORCA's key design features such as popping joints, auto-calibration, and tensioning systems that significantly reduce complexity while increasing reliability, accuracy, and robustness. We benchmark the ORCA hand across a variety of tasks, ranging from teleoperation and imitation learning to zero-shot sim-to-real reinforcement learning. Furthermore, we demonstrate its durability, withstanding more than 10,000 continuous operation cycles - equivalent to approximately 20 hours - without hardware failure, the only constraint being the duration of the experiment itself. Video is here: https://youtu.be/kUbPSYMmOds. Design files, source code, and documentation are available at https://srl.ethz.ch/orcahand.

Figures (7)

• Figure 1: (A) The ORCA hand closely mimics its human counterpart with the same form factor, a bony structure, and silicone-cast skin. The ORCA hand is 3D-printed but incorporates joints designed to pop before breaking, making it resistant to overload-induced failures while retaining the advantages of bearing pinhole joints, such as stability and simple kinematics. (A1) Just before the joint pops. (A2) Applying pressure pops the joint into place and keeps it secure. (A3) Depicts our spool system, which enables manual retention without unscrewing the spools or tendons. (B) We show that our hand can be deployed in real-world settings by running our self-resetting imitation learning policy for over 7 hours before we decided to end the experiment. (C) Our reliability test reveals our hand's robustness and the high repeatability of joint movements.
• Figure 2: Versatility of the ORCA hand: (A)-(D) Teleoperation with ROKOKO rokoko_website gloves. (A) Holding a pen (B) Using a drill, showing high dexterity. (C) Liquid pouring. (D) Grasping a cube: This picture illustrates how closely the ORCA hand resembles a human hand. (E) IL with walls and a slider for self-resetting. (F) Policies in simulation, such as rolling a ball, can be deployed zero-shot to the real world due to the ORCA hand's low joint errors.
• Figure 3: (A) Naming convention of the joints. The thumb includes an additional degree of freedom. (B) Auto-calibration: The three-step process moves all joints to their respective limits and determines a mapping between motor and joint angles without any external sensors. (C) Routing of the DIP joint. The tendons are guided around metal pins, to reduce friction and eliminate wear over time. Moreover, they are always guided through the center of rotation for straightforward control at minimal slack.
• Figure 4: (A) Joint response comparison at 0.2 Hz and 0.5 Hz: ORCA achieves accuracy and latency comparable to the LEAP hand despite its tendon-driven design. (B) Accuracy benchmarking setup using AprilTags to infer ground-truth joint angles synchronized with commands. (C) Reliability test: 2.5 hours of continuous grasping (2200+ cycles) and wrist motion (550+ cycles) without failure, overheating, or performance drop, as shown by stable motor current and temperature.
• Figure 5: Experimental setup for repeated pick & place: The cardboard serves as a fence, preventing the cube from rolling out of the testing area and enabling uninterrupted, long-duration policy deployment.