Adna is comparing 3 different RL algorithms.
Ege experimental design planned till May 2023, will be adapted based on the project extension. Ege’s exp. design and the questions will be discussed in the next meeting. Shared autonomy experiments planned to be first tested in a wire-maze task in Q2/2022.
Adna has integrated the ability to utilize human inputs to speedup the RL training time.
Franka environment will be developed in a master thesis.