Assistant professors Florian Shkurti and Igor Gilitschenski are researching new methods of vision-language models for robot manipulation technologies to improve how robotics systems perceive and safely interact with people and their environment.