Analysis of dual-sided cycling power in a virtual reality game

Virtual Reality is an upcoming topic, in which increasingly better hardware enables the realisation of realistic applications. The virtual 3D environment gives the opportunity for visual feedback. In this diploma thesis, visual feedback should be used in an application to show a user if his right and left leg power is equally distributed. Furthermore, the application should serve to improve a possible asymmetry in the leg power distribution through frequently usage. The application was developed for the use in the field of physiotherapy, especially for people with leg, knee or foot injuries and after surgery in one of these areas, as affected people usually put less strain on their injured leg and therefore a higher risk of an asymmetry consists. A persistent asymmetry in the leg power distribution can lead to premature fatigue, as well as to a worse performance and it furthermore increases the risk of injury. During the test, the user is riding a stationary bike in real world and is wearing a Head Mounted Display (HMD). In the VR environment the user is driving a pedal boat, which is steered by the measured leg power distribution of the right and left leg. The measurements are taken with a pedal-based power meter (Garmin Vector 3). In addition to the actual analysis test, where the user should only drive straight ahead, the application contains a second course, where the user drives a slalom. Different difficulty levels exist for both courses to better meet the users' needs. The bicycle analysis has a duration of ~14 -25 minutes, depending on the performance on the stationary bike. This work includes descriptions of all hardware components and the differences to existing VR bicycle applications. Further parts of this work are the explanation of the software and hardware design, details of the implementation and their visual results, as well as the description and evaluation of a user study. The performed user study showed that asymmetry in the leg power distribution can be recognized through this application. These are mostly discovered by the user himself during the usage. As every participant took the test only once, no statement can be done whether a frequently usage leads to an improvement. The analysis test was described as effective, physically demanding and entertaining.