Co-investing in soccer and care robots

Industrial robots will become more flexible thanks to soccer and care robots

Dealing with the unexpected – read: non-programmed events. For the soccer and care robots of TU Eindhoven this is a routine task – after all, you never know what an opponent or user will do next. Industrial robots, too, would benefit from increased independence and autonomy. For this reason, the university together with industrial partners have set up a pioneering research project to provide mobile industrial robots with an open world view that adapts to changes in the environment. An important aspect of this is that the robots learn to understand what they see so that they can better adapt to new circumstances.

The TU/e High Tech Systems Center, the companies Lely Industries, Vanderlande Industries, ExRobotics (ImProvia), Diversey and Rademaker combined with funding support from the Topsector HTSM (TKI allowance) are co-investing 1.5 million euros in the project, which begins in January 2018 and will run for four years. TU/e will appoint four new PhD students to the project, assisted by master students. The project goes under the name FAST: new Frontiers in Autonomous Systems Technology. The industrial partners cover a wide range of applications: agriculture, internal logistics, inspection robots, high-tech bakery equipment and professional cleaning, and the door is open for partners in other fields to join.

The project aims to develop mobile robots with increased flexibility that enables them to operate in  environments that undergo frequent changes. This has many advantages, for instance, the environment does not have to be fenced off, signaled or demarcated and specific zones or reference points will not be required. Since the usual assumption of a static world will be relaxed, and adaption of the robot to its environment will be increased, development time for a robot is expected to be reduced.

This will be realized by creating a so called ‘semantic world model’. ‘Semantic’ means that a robot can give meaning to its observations, recognize situations and objects, enabling the robot to react more appropriate to a situation. It can for example, make a distinction between a pallet with goods, a group of people or a forklift truck – the possibilities for the robot to respond appropriately are manifold. They also have to actively seek information to better assess a situation and, to ultimately, perform their task better. Examples are actively discovering the situation using multiple sensors, by getting information from the Internet of Things, or by consulting a person.

This new information adds to the robot’s world view. The robots therefore learns so that the next time the situation occurs, it can more easily deal with or prevent it. They will also increasingly operate between and with people so within the semantic world model, interaction with people is a key component.

The creation of the project is the result of an industry survey made by the TU/e High Tech Systems Center to establish shared robotics R&D questions in which the expertise of TU/e could play a role. This generated the wish to be able to make mobile robotic systems more autonomous, an area in which the university is a leader. The soccer robot and care robot teams have picked up prizes over the past decade in the RoboCup, the premier tournament for autonomous robotics.

The collaboration between the university and its industrial partners, and between partners themselves, enables experiences to be quickly shared and use cases to be added early on in the development process. This will help shorten the development turnaround and thereby consolidate the Netherlands’ position as top player in the field of mobile robotics. The project also provides a platform for students to gain relevant industry experience, something that, in the context of a growing shortage of engineering top talent, will also boost the competitiveness of the Netherlands.

The TU/e departments participating in the project are Electrical Engineering, Mechanical Engineering and Industrial Engineering & Innovation Sciences.