Typically, robots rely on human vision. Either human beings have to put items into the right places for the robots to start with, as in the typical factory, or they have to be present to manipulate the movements of the robots, as in medical robotics like the Da Vinci system.
There are systems that use light as a data source to simulate vision, but they are complex and expensive enough to have limited practical use.
There’s a new vision-guided robotics system in the UK that might change that.
A factory needed to stack bobbins of paper on palettes — a great job for a machine. However, the bobbins aren’t stacked according to number. The robot can’t simply be programmed to stack six rows of bobbins, for example, because the amount of space taken up by the bobbins varies.
So the robot was given a camera. A laser level is programmed in, and the robot can measure the height of the bobbins and of the stack they make by comparing the line in its vision with the laser line.
It’s rudimentary vision, but it’s a great example of problem-solving. This kind of thinking is what led to the development of automation, and it is still stretching the boundaries.