A preliminary model has been developed, utilizing trajectory optimization to find ‘the best robot for the task’. This model takes a given set of (randomly generated) tasks and a dynamics model with undefined motor and gearbox parameters and finds the optimal values for these parameters. However, this model lacks detailed components such as an accurate gearbox model, saturation effects and the influence of motor and gearbox selection on the dynamics model. Furthermore, no experimental validation has yet been performed, because of which the performance of the model is uncertain.
This project will focus on improving the existing model, as well as performing an experimental validation by means of building a physical prototype.