ROBOT CONTROL

Many tasks in the industry are still performed manually. This is not because of cost of labour but because industrial robots are still devoid of the unique human dexterity. Examples of such tasks are polishing of complex surfaces, fine insertion of fragile or flexible materials.

 

These tasks require  fine control of forces when in interaction. We use novel methods for force-control. Our solutions offer flexible and adaptive controllers that can vary control of forces when in contact, as humans do.

You do not  have force-sensing on your robot,. This is not a problem. We offer new methods that allow you to generate desired force-profile in the absence of explicit force-feedback.

LEARNING

The industry changes rapidly. In a few months, you may want to deploy new products that require small changes in the way polishing and insertions are done. Human workers re-learn rapidly new tasks and they transfer easily learned skills to new tasks.

We develop methods that enable robots to adapt to cnew tasks and to be rapidly trained again to a variant of the original task.

MACHINE LEARNING

We offer solutions for skilled workers to train robots on site. Our models are based on advanced machine learning techniques. 

We can learn models of the control of forces in interaction. These models can then be transferred to other tasks than those the robot is trained on.

MONITORING & PERFORMANCE ASSESSMENT

A model of skill can be used for assessing workers’ performcance at work and for monitoring skill progression.

SAFETY

Collaborative robots work next to humans. They must remain safe at all times. Our controller ensure that the robot avoids the human. In the event of a contact, the robot absorbs the contact force.

CUSTOMIZED

All controllers are customized to the client’s needs. We can use a variety of existing robotic platforms from traditional robotic manufacturers such as KUKA, ABB, UNIVERSAL ROBOTS.

We also offer advices on which robot platform to purchase depending on your needs.