Deep Learning Robotics (DLR) has developed a vision-based robotic controller that uses advanced machine-learning algorithms to enable robots to learn and replicate tasks by observing humans performing them.

The companys technology creates solutions for userrobot interaction and robot-to-robot collaboration.

The DLR device captures 3D images of a user performing the intended task on a specific object.

With the ability to differentiate between the human hand and the object, and using advanced AI machine-learning methods such as neural networks, the sequence is automatically translated into robot control language.

This intuitive task-teaching process is independent of any specific robot hardware.