When a human being walks around he constantly monitors the environment and takes his steps accordingly. When a robot has to walk around it has usually got pre-fed data on the path and can not always monitor changes in real time. Robotic engineers need to ensure that the robots that they design don’t go crashing into things in their way.

This design feature is one that has been constantly evolving with each new generation of robots. With the Simultaneous Localization and Mapping (SLAM) the scientists at MIT have given computers a better chance of not banging into things in their path. The system uses a camera to take images constantly as the robot moves and compare each picture with the last one. They started out using a laser, but since that was cost prohibitive, they switched to relatively cheaper digital cameras.

The new obstacles that enter the field of motion are navigated and the distance from the last point is noted. This constant ongoing evaluation of the unknown environment allows the robot to move comfortably in areas that is has never visited before. With an addition of a voice directions feature the system also has a possible use for blind people to navigate through new paths on their own as well. Now that would be a really useful science project.