Comparing the Effects of Admittance and Error-based Control of a Trunk Rehabilitation Robot on the User’s Balance

Trunk training is an important part of the rehabilitation of people suffering from the physical effects of brain lesions caused by diseases such as stroke, as it can help improve their static and dynamic balance, while also helping to improve gait performance. In order to facilitate trunk rehabilitation and reduce therapist workload, we have developed a trunk rehabilitation robot (TRR) that can provide tunable quantitative training with the option to include sensory augmentation through various bio-feedback, in order to implement a large variety of trunk training protocols. In this paper, we have developed an error-based controller to generate the unstable seat condition using the TRR. In order to verify the feasibility of this controller, we have carried out tests with 20 young healthy subjects and compared its performance with the existing admittance controller in terms of seat movement parameters and center of pressure and trunk movement based balance parameters. The results show that although there is no statistically significant difference between the controllers in terms of the balance measures, use of the error-based controller results in lesser amount of seat movement, which may make it more comfortable for the user. However, the participants had a difference of opinion about which controller they felt to be more comfortable and easier to use.