# Ability Hand by PSYONIC Canonical page: https://technooptimist.io/neurotech/ability-hand Frontier: [Neurotech](https://technooptimist.io/neurotech.md) · Field: Bionic limbs & muscle control · Company: [PSYONIC](https://technooptimist.io/companies/psyonic.md) Stage: Deploy (4 of 5: Research → Proto → Pilot → Deploy → Scaling) Updated: last update 2026-06-16, facts checked 2026-10-11 Goal: Give people missing a hand a fast, tough prosthesis they can feel through, at a price Medicare covers, and use the same hand to collect human grasp data that teaches robots delicate handling. Status: Fitted to about 300 amputees and sold to robot labs; ABB and NVIDIA tie-ups in 2026 Next milestone: Reach >99% operational reliability for robot use ## Milestones | Status | Milestone | Date | Slip | | --- | --- | --- | --- | | Done | 3D-printed prototype fitted in Ecuador | 2014 | | | Done | Wins University of Illinois Cozad venture competition | 2015 | | | Done | Ability Hand released with multi-touch feedback | Sep 2021 | | | Done | Native asset in NVIDIA Isaac Lab (GTC) | 17 Mar 2026 | | | Done | ABB GoFa cobot collaboration | 16 Jun 2026 | | | Current | Reach >99% operational reliability for robot use | | | ## Most important updates - **16 Jun 2026**: ABB Robotics and PSYONIC pair the Ability Hand with the GoFa cobot to train robots on human grasping data. ABB Robotics is working with PSYONIC to study how touch and motion data from people using the Ability Hand can teach robots delicate, variable tasks that are hard to automate. ([our story](https://technooptimist.io/stories/abb-robotics-and-psyonic-pair-the-ability-hand-with-the-gofa-cobot-to-ff534635.md) · [original source](https://www.cityam.com/abb-robotics-and-psyonic-use-human-generated-data-to-advance-robotic-dexterity/)) - **17 Mar 2026**: PSYONIC's Ability Hand becomes a native asset in NVIDIA Isaac Lab for robot learning. At NVIDIA GTC, PSYONIC and NVIDIA announced that the Ability Hand is now built into Isaac Lab, so researchers can simulate, train and test AI grasping policies on it. ([our story](https://technooptimist.io/stories/psyonic-s-ability-hand-becomes-a-native-asset-in-nvidia-isaac-lab-for-12041c32.md) · [original source](https://www.humanoidsdaily.com/news/from-prosthetics-to-pixels-psyonic-and-nvidia-bridge-the-robotics-data-gap-with-real-to-real-transfer)) - **2 Sep 2021**: PSYONIC releases the Ability Hand, a bionic hand with multi-touch sensory feedback. PSYONIC launched the Ability Hand, a bionic hand that gives users touch feedback and closes in 200 milliseconds. ([our story](https://technooptimist.io/stories/psyonic-releases-the-ability-hand-a-bionic-hand-with-multi-touch-688bc41c.md) · [original source](https://rehabpub.com/orthotics-prosthetics/prosthetics/psyonic-launches-the-ability-hand-a-bionic-hand-you-can-arm-wrestle-with/)) ## Current obstacles - **Slow grip selection**: Picking among 32 grips with a few muscle signals takes practice and seconds per switch, far from a natural hand's effortless control. - **Proving durability claims**: Speed and robustness claims are mostly company-reported; independent long-term field data on failure rates is limited. ## Physics limits - **Skin electrodes see a blurred signal**: Surface electrodes read the summed activity of many muscles through skin and fat; sweat and socket shift change the signal, capping how many independent movements can be decoded. - **Vibration is not real touch**: A buzz on the arm tells you force, not texture, slip or which finger touched; true sensation needs nerve stimulation inside the body. ## How it works ### Reading intent: Muscle signals pick the grip Electrodes in the socket read forearm muscle activity; contracting a muscle opens or closes the hand, and set patterns switch between grips. ### Moving: Five powered fingers, rotating thumb All five fingers flex and extend and the thumb rotates electrically, giving 32 grip patterns and closing in 200 ms. ### Feeling: Pressure sensors drive a vibration Fingertip sensors measure contact force and a vibration on the arm reports it, so users can grade grip strength without looking. ### Teaching robots: Same hand on people and robots Because robots use identical hardware, a human user's movement, contact and force data can be replayed to train robot grasping policies. ## Spec sheet | Spec | Ability Hand | Provenance | | --- | --- | --- | | Weight | 490 g | reported | | Closing speed | 200 ms | reported | | Grip patterns | 32 (19 predefined) | reported | | Battery life | 6-8 h per charge, ~1 h USB-C charge | reported | | Water resistance | IP64 (splash resistant) | reported | | Amputee users | About 300 | reported | Provenance: reported = stated by the company; estimated = our estimate; sample = a sample figure. ## Papers and demos - **17 Mar 2026** (demo): [Real-to-real transfer: a prosthesis user's pipetting data trains robot arms, dogs and humanoids](https://www.humanoidsdaily.com/news/from-prosthetics-to-pixels-psyonic-and-nvidia-bridge-the-robotics-data-gap-with-real-to-real-transfer). Shows one hand can collect human grasp data and replay it on robots. ## About PSYONIC PSYONIC is a San Diego company that builds the Ability Hand, a fast, impact-tolerant bionic hand that gives amputees touch feedback. The same hand is now sold to robotics labs, so data from human users can teach robots to grasp. - Born from a 3D-printed hand Aadeel Akhtar fitted in Ecuador in 2014 as a University of Illinois grad student. - Ability Hand is reimbursable by Medicare, which the company says widens access from about 10% to 75% of users. - 2026: the hand became a native asset in NVIDIA Isaac Lab (March) and was paired with ABB's GoFa cobot (June). - Founded: 2015 - Headquarters: US - Status: private - Website: https://www.psyonic.io - People: Aadeel Akhtar (Founder and CEO) - Partners, customers, investors: ABB Robotics (partner); NVIDIA (partner); National Science Foundation (investor) Company page: https://technooptimist.io/companies/psyonic.md ## Related programs Everything in Neurotech: https://technooptimist.io/neurotech.md --- The Techno Optimist tracks every frontier of technology, program by program: https://technooptimist.io/ (a Markdown version of any page: add .md to its URL; site map for agents: https://technooptimist.io/llms.txt). Data API (JSON, realtime stream): https://technooptimist.io/developers. People read 5 pages free, then Reader $10/month; these Markdown pages are open to agents.