Summary

Advanced human robotic interaction has for a long time been a topic mainly for science fiction and academic research. The last few years advancing AI technologies have however bridged the gap between science fiction and reality, and a future where we will meet and interact with robots almost every day doesn’t seem far away. But while AI models are getting smarter in conversation and reasoning their moving patterns and facial expressions can still be perceived as stiff and static. Genetic algorithms might however offer ways of searching for interaction patterns that appear pleasant to humans. In this thesis, students will therefore explore how genetic algorithms can be used to find such patterns. 

To their help they will have the vehicle Will-e which is small electric vehicle at Semcon office. Equipped with lidar, camera and radar, he has the possibility of detecting people and other objects in his vicinity. He also...