TU Berlin
Department of Computer Science
 TU Berlin

Institute for Technical Computer Science
Real Time Systems & Robotics

Powered Hand Exoskeleton

The powered hand exoskeleton for rehabilitation purposes
with multiple actuated degrees of freedom

First Version of Hand-Exoskeleton


The powered hand exoskeleton is an orthotic device that is worn by a human. Individual joints of the human hand can be moved. The actuation is performed by electric motors which transmit forces to the joints through Bowden cables. A real-time controller calculates the necessary control signals for the motor controllers based on the measured forces and joint angles. A separate interface computer allows the therapist to change the control modes, define new exercises, and supervise the rehabilitation.

Following features are integrated to support movement and diagnosis:
  • 20 active degrees of freedom (four for each finger)
  • Joint angle sensors (Hall sensor and encoders)
  • Force sensors at finger tips
  • Up to 16 EMG Sensors

Other Possible Applications

The device was developed with the focus on the rehabilitation of hand injuries. Other possible application of hand exoskeleton devices are:

  • Haptic feedback for virtual reality
  • Telemanipulation
  • Assistive device

Control Modes

For the support of rehabilitation a position control based on sliding mode control was implemented. Force based control is also supported by using force sensor data to generate a trajectory for the position control loop. Force control can be used to simulate a minimal friction of the environment. The hand exoskeleton then basically follows the human movements. By increasing simulated friction the human must use more force to move the hand exoskeleton which is useful to strengthen the muscles. By mixing of position and force control guided movements with human interaction become possible.

The hand exoskeleton therefore allows to support rehabilitation by following exercises:
  • Movement with no human interaction (joint mobilisation)
  • Force-based control (force enhancement, muscle training)
  • Mixed force and position control (e.g. guided movement)
  • Force measurement (e.g. for assessment)
This research is funded by the German Research Foundation (DFG).

Contact persons:
Dipl.-Ing. Andreas Wege (E-Mail)
Last update 06.09.2007 by Andreas Wege

Real Time Systems & Robotics