Industrial robots have been around for some 50+ years with the first officially recognised device being created in 1954. It wasn’t until 1960 though that the first robot company was formed and of course all robot guidance was by control systems with known coordinates whose accuracy was highly critical with any deviation in either product variation or position errors causing major headaches for the end customer. This state of affairs continued for some time - until at least 1983 when the first commercial vision systems became available.
Machine vision has had a long period of development since it became a viable technology and has been very widely deployed within many industries over time. Vision brings benefits to many sectors but robotics is one of the big winners for many reasons. One key point is the fact that it can be seen as an enabling technology in robot guidance. Robots are good at repetitive tasks but they cannot accommodate changing parameters, so when a location changes or a product is not where it is suppose to be, the robot system will fail.
Machine vision enables robots to ‘see’ an object and calculate it’s X and Y position (and X, in relation to the robot picking arm. But it also enables the robot to ‘see’ the correct placing position.
More recently, robot guidance has evolved with 3D machine vision. So a third co-ordinate is also available, typically the height of an object.
There is a steadily increasing range of smart cameras to add to the list of established image sensors and software packages, there are robot guidance systems for any application. With the advent of low cost multi-core processors more can be done