Joshua Van Bockel
Contact information
Location: Building 193, Level 4, Desk 410.01
Thesis title
Integrating emotional intelligence into Reinforcement Learning for enhancing organic gait and behaviour in a quadrupedal robot
Research overview
This research aims to improve the interaction between humans and robots by studying how human emotional intelligence models can be incorporated into the Reinforcement Learning framework.
Since robots are increasingly prevalent in the workforce and in the home, humans and robots will be interacting, communicating and working together much more often. This research aims to make that interaction more seamless and natural, so robots are easier and more enjoyable to be with.
Particularly, this research will involve a quadrupedal robot dog, the Unitree Go1 which provides the hardware to test the experimental emotional artificial intelligence, and realistic movement and behaviour can be assessed.
Supervisors
Qualifications
B. Sc, University of Melbourne; M. Eng Mechatronics, University of Melbourne