Home Research Project Details 3a - Self-learning, predicting control methods for body supporting systems, especially orthopaedic devices for the lower extremities
Personal tools

3a - Self-learning, predicting control methods for body supporting systems, especially orthopaedic devices for the lower extremities

Florentin Wörgötter and Otto Bock HealthCare GmbH

The goal of this project is to develop a novel type of orthosis, which is able to quickly adapt to new (walking-) situations and which can, through the use of advanced sensors, also to some degree anticipate actions planned by the patient (e.g. sitting down, walking upstairs, etc.). To this end, novel adaptive neuronal control methods will be used similar to networks  successfully employed in robot control (“RunBot”).

For the stabilization of a patient’s musculoskeletal system most often KAFO-type (knee-ankle-foot) orthoses with locked-knee joints (released manually by the patient, e.g. for sitting down) are used. Recently, mechatronic orthotic joints appeared on the market. Most of them are restricted by a gait depending switch of the joints based on mechanic non-adaptive switches (Irby et al 2007, Sabelis et al 2007, Bernhardt et al 2007). Common disturbances (floor unevenness, obstacles, ramps) can not be mastered in a satisfactory way. This strongly restricts acceptance and possible medical indications of such systems. Fully actuated orthoses are practically infeasible due to their high energy use. As a consequence, novel approaches include active elements into the orthosis but mostly in the form of small actuators, which do not directly act on the movement, but, instead, adjust the compliance only (!) of the heel switching of the device, relying on sensor information (e.g. Fior&Gentz, “NeuroTronic”; Otto Bock Healthcare, “Sensor-Walk”). Currently, the efficient control of such actuators remains a very difficult problem as it requires advanced sensor signal analysis and fusion methods as well as controllers, which are compliant and “lenient enough” not trying to enforce unwanted actions on the user.


Fig. SP3a) Left, „E-MAG Active" orthosis by Otto Bock Healthcare. Right, Run-Bot's control architecture. Sensors (G, ground contact, A hip angle IR infra red, AS accelerometer) and elements S,N are modeled as rate coded adaptive neurons. Different levels of control with different degrees of autonomy similar to biomechanical, spinal and central nervous control are implemented. The walking process (Environment) retriggers sensor inputs. Neural elements are adaptive and can learn via a learning control circuit (not shown).

Belongs to Group(s):
Otto Bock HealthCare GmbH

Is part of  Section 3 

Members working within this Project:
Manoonpong, Poramate 
Wörgötter, Florentin 
Kroll-Orywahl, Olaf 

Selected Publication(s):

Braun, J, Wörgötter, F, and Manoonpong, P (2014).
Internal Models Support Specic Gaits in Orthotic Devices
WSPC Proceedings:1--8.

Kuhlemann, I, Braun, J, Wörgötter, F, and Manoonpong, P (2014).
Comparing Arc-shaped Feet and Rigid Ankles with Flat Feet and Compliant Ankles for a Dynamic Walker
WSPC Proceedings:1--8.

Liu, G, Wörgötter, F, and Markelic, I (2013).
Stochastic Lane Shape Estimation Using Local Image Descriptors
IEEE Transactions on Intelligent Transportation Systems 14(1):13-21.

Kulvicius, T, KeJun, N, Tamosiunaite, M, and Wörgötter, F (2012).
Joining Movement Sequences: Modified Dynamic Movement Primitives for Robotics Applications Exemplified on Handwriting
IEEE Transactions on Robotics 28(1):145-157 doi:10.1109/TRO.2011.2163863.

Liu, G, Wörgötter, F, and Markelic, I (2012).
Square-Root Sigma-Point Information Filtering
IEEE Transactions on Automatic Control 57(11):2945-2950.

Kulvicius, T, KeJun, N, Tamosiunaite, M, and Wörgötter, F (2011).
Accurate position and velocity control for trajectories based on dynamic movement primitives
IEEE International Conference on Robotics and Automation (ICRA):5006-5011 doi:10.1109/ICRA.2011.5979668.

Kulvicius, T, KeJun, N, Tamosiunaite, M, and Wörgötter, F (2011).
Modified dynamic movement primitives for joining movement sequences
IEEE International Conference on Robotics and Automation (ICRA):2275-2280 doi:10.1109/ICRA.2011.5979716.