MIRA University of Twente
Department of Biomechanical Engineering

Design of a SEA for LOPES (C. Lagoda)


Claude Lagoda


Biomedical Engineering (TU Delft / UTwente)




Dr. ir. Herman van der Kooij (project coordinator)

Dr.ir. Alfred C. Schouten (supervisor TU Delft)

Starting date:


Claude Lagoda


Given in Delft, colloquium expected in June 2009


Design of a new series elastic actuator (SEA) for the gait rehabilitation robot LOPES


Stroke (also known as Cerebro Vascular Accident) is the leading cause of death in Western civilization. In many cases it leads to serious brain damage which is the reason for human motor control disabilities amongst others. If stroke survivors suffer from hemiplegia (paralysis affecting one side of the body), rehabilitation is needed.

The LOwer Extremity Powered ExoSkeleton (LOPES) project focuses on gait training and is used in combination with a treadmill to rehabilitate hemiplegic patients. The major goals of LOPES are to allow more efficient gait training by assisting as needed and to reduce the physical workload of physiotherapists during training. Two different types of control shall be realized: “Patient in charge” and “Robot in charge” mode. In order to realize these modes there is a need to make use of impedance control instead of admittance control.

Impedance control demands for lightweight, low friction construction and a low impedance actuator to reduce intrinsic mechanical impedance of the exoskeleton.

Admittance control needs high positioning bandwidth and thus high power actuation and rigid construction. The measurement of low stiffness, occurring during free motion, is low and the controller gets unstable.

The demands for an impedance controlled actuator will be consequently:


Have enough power and bandwidth.


Facilitate accurate force control.


Allow for an impedance control range from imperceptible to stiff.


Add as little mass as possible to the exoskeleton.

A solution could be the use of a series elastic actuator (SEA). This type of actuator is in general a series connection of a DC-motor and a compliant element, normally a spring. The advantage is that the spring decouples the dynamics of the actuator from the load. This results in increased backdrivability of the actuator, reduced friction and backlash, filtering of non-idealities in transmission drives from the force output and absorption of shock loadings from the load which protects the motor. The disadvantage is a reduced modulating frequency for large output forces. This limitation in force bandwidth decreases positioning-speed and positioning-accuracy. A trade-off between maximally achievable output force and speed bandwidth has to be found.


The goal of the thesis project is to find an alternative way of actuating the joints of the exoskeleton. Replacing the Bowden cables by another mechanism has been found unsatisfactory and thus a solution mounted on the joint has been chosen.

A new SEA will be developed and tested. The SEA has to be relatively lightweight (mass of
2 kg), compact and powerful (100 Nm peak torque at 5 rad/s).

Furthermore the exoskeleton frame will have to be adapted to the new actuation system.



Different concepts of the SEA will be developed and the best one will be chosen.


A special spring design will be developed, simulated and tested in reality.


A prototype of the SEA will be built, tested and its properties compared to existing solutions.


The exoskeleton frame will be adapted to the new SEA (only as 3D model).

Further information

Master thesis project is in the stage of building the prototype. (as at November 2008).