Human-robot collaborative learning for ultralight electric utility vehicles
Facts
- Contact person:
- Paul Davidsson
- Financer:
-
- Formas
- Responsible at MaU:
- Paul Davidsson
- Collaborators :
-
- LEVTEK SWEDEN AB
- PostNord AB
- Retta AB
- MyoFysio Helsingborg AB
- Time frame:
- 01 December 2024 - 28 February 2027
- Faculty/department:
- Research environment :
- Research subject:
About the project
Collaborative robots are designed to work alongside humans in a shared workspace, unlike traditional industrial robots that mainly operate in isolated environments. While there exists autonomous mobile robots designed for collaborative usage, there are no commercially available collaborative vehicles that a human user can both drive/ride and work alongside in a practical and safe way. However, there is great potential for ultralight electric vehicles with these features. The compact and open form factors enable easy switching between driving/riding and working alongside the vehicle. The lighter weight and slower operating speeds makes them safer and usable in a wide range of environments.
Objectives
In the proposed project, the objective is to develop and implement human-robot collaborative learning for ultralight utility vehicles with autonomous capabilities being developed by LEVTEK SWEDEN AB. We will use a unique approach that combines end-to-end imitation learning with interactive machine learning (IML) developed at Malmö University. To make sure it is both usable and valuable, physiotherapists from MyoFysio AB and end-users at PostNord AB and Retta AB are included in the project team. The general goal is to make the transport of both people and cargo more efficient and sustainable for both the environment and users. Another goal in the proposed project is to implement and validate research results regarding IML in realistic settings.