Artificial Intelligence and Terrain Handling of a Hexapod Robot

The focus of this master’s thesis has been getting a six-legged robot to autonomously navigate around a room with obstacles. To make this possible a time-of-flight camera has been implemented in order to map the environment. Two force sensing resistors were also mounted on the front legs to sense if the legs are in contact with the ground. The information from the sensors together with a path planning algorithm and new leg trajectory calculations have resulted in new abilities for the robot. During the project comparisons between different hardware, tools and algorithms have been made in order to choose the ones fitting the requirements best. Tests have also been made, both on the robot and in simulations to investigate how well the models work. The results are path planning algorithms and a terrain handling feature that works very well in simulations and satisfyingly in the real world. One of the main bottlenecks during the project has been the lack of computational power of the hardware. With a stronger processor, the algorithms would work more efficiently and could be developed further to increase the precision of the environment map and the path planning algorithms