Video Demonstrations

The following videos show our system in action. Our system can procedurally map the environment as well as navigate around obstacles to avoid them. This is displayed on a graphical interface for the user to see.

SLAM Demo

This is our very first prototype for our system. We simply carried a LiDAR sensor around a room while our system performed SLAM. As we move, the system is able to track our movement and plot the path that we take. Then the LiDAR data is plotted as well: Solid points which the LiDAR found are marked in black, and open space is plotted in white. By the end, the building's hallways, rooms, and windows are visible. You may notice some flaws in the resulting map--there are rooms and walls in places there shouldn't be. This is because of human error with handling the LiDAR, forcing the system to adjust as time goes on. We finished this prototype by the end of Spring, and we moved to expand on it in the coming months.

Navigation Demo

Our second prototype of our system includes both SLAM and navigation functionality. The robot is given a point which it will attempt to move to. Our system chooses the best route there based on how short the route is and how close it must venture towards obstacles. This is determined mathematically by a cost map, which is shown in the display. Obstacles are surrounded by a color gradient which represents distance away from an object, with closer distances being more dangerous and therefore having higher cost. The system will take the lowest cost route possible. If the robot must move through a high-cost region, the robot will move very slowly for safety purposes.