Skip to content

Use with AI

Use with AI

Tools and hands ​

The object and process determine the tool: repeatable gripping, adaptive enveloping or coordinated finger motion. Each arm has a tool profile with mechanics, power, driver and feedback.

Choose the contact ​

Control and contact ​

Commands, contact and feedback define the tooling concept. For hands, DoA counts independent actuators; DoF describes motion freedoms. Coupling and underactuation distribute inputs across the fingers. Teleoperation maps human-hand motion to the actual device axes.

Compare tooling concepts

Tool example SCHUNK EGK 25

Command
Opening width / grip
Contact
Defined finger surfaces
Feedback
Position / grip status

The selected tool profile defines jaw stroke, force limits and part detection.

7 / 7 hands · grouped by control concept

Manufacturer product image of the Prensilia Mia Hand R&D variant
Mia Hand · R&D

Mia Hand

3 actuators · 4 DoF

Details & sources
Application
Adaptive grasping of everyday objects and tools with a compact command space.
Control concept
Middle, ring and little fingers are coupled. Industrial: seven preset grasps with position/speed control.
Interfaces
Industrial: RS-485, 24 V. R&D: UART over USB with manufacturer ROS 1/ROS 2 packages.
Feedback & integration
Position, motor current and limit switches; optional force sensors. Force control is announced.
Manufacturer image of Adam’s Hand, AH02 image series
Adam’s Hand · AH02

Adam’s Hand / Plus

Plus: 3 motors · 10 driven DoF

Details & sources
Application
Enveloping grasps that adapt mechanically to the object; useful when individual finger trajectories are not the task.
Control concept
Plus: one motor with differentials drives four fingers; two motors control thumb flexion and rotation. The ten driven DoF are mechanically coupled; the classic prosthetic model uses two motors.
Interfaces
Bluetooth, Wi-Fi and SDK. RS-485, CAN, Ethernet and native ROS 2 are announced.
Feedback & integration
Manufacturer lists current, position, speed, IMU and motor torque. Force/temperature sensing and slip detection are announced.
Manufacturer product image of the INSPIRE ROBOTS RH56F1 hand

RH56F1

6 DoF · 12 joints

Details & sources
Application
Anthropomorphic grasping with a small control vector; a concrete EtherCAT option for peripheral integration.
Control concept
Six controlled DoF coordinate twelve joints. Grasp and teleoperation targets map to six device axes.
Interfaces
E2: EtherCAT + RS-485; E4: EtherCAT + CAN FD. Supply: 24–48 V; left/right hand and sensing follow the variant.
Feedback & integration
T1 variants add eight tactile sensors according to the manufacturer’s variant table.
Palm view of the Alt-Bionics Surge hand from the manufacturer’s v1.3 datasheet

Surge

6 powered DoF · modular fingers

Details & sources
Application
Grasping and glove-based teleoperation with individually replaceable finger modules.
Control concept
Six powered DoF with MANUS glove compatibility. Retargeting maps glove motion to the hand’s axes and limits.
Interfaces
RS-485; 48 V nominal (datasheet v1.3). Protocol and firmware are selected for the hand variant.
Feedback & integration
Optional tactile force sensing: 80 individually addressable taxels per finger. Taxels are sensing elements, independent of motion DoF.
Manufacturer product image of the white VINCENTevolution5 hand

VINCENTevolution5

6 motors · 6 independent DoF · 11 joints

Details & sources
Application
Human-shaped handling for service and humanoid robotics; demonstrated by the manufacturer on KIT’s ARMAR-7.
Control concept
Six motors control six independent DoF across eleven joints. Interface and software are adapted for robotics.
Interfaces
Robotics interface, power supply and SDK are specified for the selected robotics variant.
Feedback & integration
Telemetry fields and update rates are specified with the selected robotics interface.
Manufacturer image of the ARTUS Lite hand released in 2024

ARTUS Lite

20 DoF · 16 actuated + 4 underactuated

Details & sources
Application
Fine manipulation, finger coordination and learning tasks that require many independent joint targets.
Control concept
Lite API: sixteen joint targets and four underactuated DoF. The hand model defines joint order, limits and retargeting.
Interfaces
ARTUS family Python API: RS-485/Modbus RTU or TCP. Transport and API version follow the model’s compatibility profile.
Feedback & integration
State reads, calibration and fault handling follow the selected hand, firmware and API profile.
AEON Robotics HandEffector holding a tomato in the manufacturer’s product image

HandEffector

11 / 14 QDD actuators · 11–19 DoF

Details & sources
Application
Contact-sensitive grasping with joint torque information and perception close to the object.
Control concept
Quasi-direct-drive actuators with joint torque sensing. Actuator count, passive joints and finger abduction follow the configuration.
Interfaces
RS-485 and USB, 24 V. Command modes and update rates are specified for the selected configuration.
Feedback & integration
Joint position, velocity and torque; integrated 5 MP RGB camera, ToF depth sensor and ARM processor with AI accelerator.

One profile per tool ​

LayerProfile content
Mechanics & powerLeft/right variant, adapter, tool frame, mass/centre of gravity, voltage and peak current
DriverBus and firmware; grasp presets, finger-group or joint targets; supported command modes and limits
FeedbackPosition/current, joint torque, fingertip force or tactile arrays; timestamps and calibration
BehaviourGrasp confirmation, communication timeout, release/hold, reset and recovery

Contact at the wrist ​

A six-axis wrist sensor measures the resultant force and moment. Tactile elements measure local finger or palm contacts. Select one or both wrists for the process and update the tool profile when changing tools.

Continue with force/torque sensing, Control and interfaces and the hardware ecosystem.