An infrared proximity sensor on a robot can give it â€˜eyesightâ€™ to make sense of the real world. Cory Barton has used an inexpensive Sharp distance sensor, an Arduino and the open source Processing language to build his robotics project. The idea is to use the infrared sensors mounted on the robot to detect surrounding objects. Then his program processes the readings to get a virtual sense of what the sensors are seeing.
The hardware for the vision system is a Sharp GP2Y0A21YK0F IR distance sensor mounted on a standard hobby servo. It interfaces with an Arduino Diecimila board connected to a computer. The disadvantage of the inexpensive (less than $10) Sharp IR sensor is its narrow field of focus. It can detect objects only in the range of 10-80cm, so Cory has his mounted on the servo which allows for 180 degree pan – the analog voltage measurements indicating the distance. Sensor panning with multiple readings gives a better approximation of the layout.
A programmable Arduino IDE controls the servo-sensor combine and feeds the readings to a small program written in the Programming language on the computer. The program processes the data into a pie-chart which helps to visualize the layout around the robot. This inexpensive project goes into the basics of how a visually sensing robot can be built. AI buffs can check out some more of our cool projects in the Robots section.