The research group Intelligent Robotics figures out how robots can be designed to interact in the best way with humans, other robots, and the environment.
Our research follows two main tracks: Field Robotics and Human-Robot Interaction. In the area of Field Robotics, we have done a lot of research on automation in forestry, in particular development of autonomous forest machines. In the FP7 project CROPS we developed robots for harvesting and spraying of fruit and vegetable crops, and continued in the H2020 project SWEEPER, where the consortium with joint efforts succeeded in developing the first sweet-pepper harvesting robot demonstrated in green house environments.
Our second main track of research is Human-Robot Interaction, and we have coordinates several EU projects in that area: INTRO (FP7), SOCRATES (H2020). We initially focused on cognitive architectures and robot learning, primarily Learning From Demonstration (LFD) enhanced by HRI. We currently focus on Understandability in robotics, i.e. how advanced robots can communicate the content of their artificial minds to interacting humans. For this we develop techniques for natural language understanding, dialogue management, intention recognition, and task planning.
An additional SEK 20 million is invested by the Government in AI competence.
PhD students attended a two-day course in cloud computing.
As Cinnober becomes part of Nasdaq, Umeå will have a new global brand in town.
Computers represents the semantic meaning of a sentence by graphs. These graphs can be made more explicit.
"We will integrate ethical and societal concerns into the design of AI systems."