While it is easy to give a robot the sense of touch and a concept of distance, having robots that can see and process their surroundings would give us a huge advantage with the robots we build and compete with. Every match starts with an autonomous period, where robots must operate with only pre-programmed instructions. While most teams program their robots to do a set task without paying much attention to its surroundings, a robot-mounted camera would give us the opportunity to let our robot “see” motion around it and react accordingly, preventing robot collisions and increasing our score. For the rest of the match, a camera might be used to relay information to drivers about what is happening around an obstruction (which are present in some years’ games) or better yet to automate certain tasks that the drivers might otherwise have to do themselves. For example, following a brightly colored game element or the power of a shooter based on an accurate distance to a field element or scoring area.
Although a camcorder type camera can be used, they are too large for our application. We chose to use the imaging device from a IP based security camera and take advantage of its video streaming capabilities which can be applied to all of the above mention scenarios. You can find the camera we are using here.
In addition to on-robot related applications, we could also review match videos to aid in driver training. While not mounted to the robot, we would be able to use the camera to live-stream our build meetings to the general public and for some fun projects (an idea one member of the team had was superimposing pig ears on people as they visit our pits at competition). The students on our team are extremely creative, we will take whatever equipment we have access to and turn it into something useful and fun.