Machine learning and control for advanced planetary rover operations

点击翻译按钮获取中文摘要

This research program studies space robots, and their ability to traverse challenging planetary terrains. As robotic space exploration targets more and more extreme terrains on the Moon and Mars an accurate understanding of how robots can recognize and predict risks improves chances of success. The entrapment of the Mars Exploration Rover Spirit in soft soil and the tears and punctures in the Mars Science Laboratory Curiosity rover's wheels demonstrate some of the current challenges of roving on Mars that this research will address.The goal of this research is to make advances in artificial intelligence and optimal control to produce better-informed risk analysis and energy-efficiency for rovers. In addition to addressing current challenges, this research also opens new opportunities for space exploration. Advanced rovers able to drive on steeper sandy slopes, for example, open exploration to a significant number of new craters and mountains. Efficient planning of robot routes can get more out of power-starved space missions.These areas of research will help maintain Canada's global leadership in space robotics, which provides the economy with a supply of high-tech jobs and inspires the next generation of Canadians to pursue science and engineering. The development of AI technology for space robotics will lead to significant socio-economic impact in Canada by training highly qualified personnel, attracting and retaining outstanding talent in Canada, further strengthening the AI ecosystem in Canada, creating new employment or entrepreneurship opportunities, and creating new revenue streams for the use of data in space exploration space industry.