For cost and quality reasons or due to a lack of skilled staff, more and more manufacturing companies are eliminating manual tasks. But even in highly automated industries such as the automotive or electronics sector, there are still plenty of manual workstations. The reason: Variances make production steps unpredictable and automation impossible or very expensive.

The handling of cables is one of these tasks. Gripping, guiding as well as plugging and unplugging cables is not possible for conventional automation solutions. Cables – whether cable harnesses, power, data, or others – are characterized by a high degree of flexibility and form instability: They are bendable, change their shape as soon as they are moved, and must be precisely positioned and guided with the right force to be connected correctly. In production, moreover, the position of cables is not always predictable: even hanging from the air, cables must be gripped or inserted.

For us, it’s easy: We see where the cable is, how it bends and can continuously adapt our movement to the target when gripping and guiding. To be able to deal with variances in production, robots need one thing above all: hand-eye coordination.

Thanks to AI, robots now get just that: The AI-based controller MIRAI from Micropsi Industries enables robots to perceive their environment and react to it in real time. Thanks to machine learning and a camera, MIRAI-controlled robots can robustly handle variances. As a result, for the first time, they learn to handle unstable objects such as cables: grasping, guiding, plugging and unplugging them, even when they hang freely in space or the position of the socket changes.

The system is trained in just a few minutes or hours, can be adapted for other activities in a very short time, and proves to be a true multifunctional tool: From leak testing to picking reflective workpieces to assembly despite variances – AI and robotics are tapping into limitless automation in diverse industries.