New demands on renewable energy in transportation and the development of autonomous vehicles have made vehicle companies predict that their future number of variants will increase. Until now truck manufacturing has needed few assembly lines to produce all the companyís models. When hybrid, fully electric, fuel cell-based powertrains and different grades of autonomous vehicles get introduced, the lines will become crowded, and the different parts will be hard to fit next to the production. The ViMCoR project focuses on developing ATRs (Autonomous Transport Robots), that will support the line with parts from remote places and order new material when the AI system detects that more is needed. It is also building an AI system that will improve collaborative robotsí abilities to work side by side with humans at the assembly line.
– We are dealing with future flexible manufacturing where people and robots work together in a unified environment on equal terms. Todayí standards have many barriers regarding efficiency and business case when it comes to safety and arenít following the development of technology and acceptance of these new technologies in sociaty. We intend to use the human flexibility as a resource in the safety loop. We have demands on both humans and robots. ViMCoR is about utilizing this flexibility, says Per-Lage Götwall, Volvo's representative in the project.
By training the collaborative robots in working side by side with human workers the work process will become smoother. When people build trucks, they have a certain movement pattern. The AI system will learn to predict what the human is going to do next and reacts accordingly. If the project manages to develop a safe and useful algorithm this will eventually provide a business case and decrease the need of building new assembly lines for every new model.
Also, the ATRs are trained to predict human movement and being able to reroute its way when obstacles appear. With multiple cameras equipped with computer vision placed in the factory ceiling an AI system will be able to monitor the whole environment and predict dangers or obstacles for the ATRs to find other ways around.
– We are trying to find a good balance between what people are good in doing and use robots to what they do the best. In our vision we see a need for perhaps a thousand ATRs that deliver material inside a factory and it is important to reduce the cost for every ATR, says Knut Åkesson project leader at Chalmers.
The project went in 2020 from constructing methods to set the scene, and will continue with eight master students building the AI system.