Discussion

These results were expected as it was our end goal to have a fully functioning robot that was able to utilize both facial recognition and pathfinding to get to its customers. If we had more time, we would have definitely tried to see how the robot would have done if it had a larger base of customers and developed a deep neural network for more accurate facial recognition. The limitations of our findings include our assumption that the robot would have a directly clear path to the customer. In an actual restaurant scenario, there would be customers, objects and staff blocking the robots' path. Although our robot is able to successfully move past static obstacles, we don't know its reaction to a moving object. Testing that and coding in its reactions would be an excellent way to expand on our project.