AI and Robotics Lab cooperate on innovative cyber-physical systems

Halldor Arnason started his PhD project this summer and is going to apply and build on machine learning techniques for intelligent control related to manufacturing. Work on different forms of reinforcement learning has been going on for several years, especially with regard to stategic control of batteries and other assets in an electricity system.  MSc graduate, Jostein Danielsen investigated last spring new possibilities related to camera settings and artificial vision related to correct component detection and robot handling.  

See more here.

Faraz Kanafi Safarpour, also a MSc graduate, has worked together with Halldor on cyber-physical security that has now materialized in a new scientific paper. PhD students Andreas Dyrøy Jansson, based in Narvik and Yigit Can Dundar and his main supervisor, Professor Anne Håkansson in Tromsø have all worked on multi-agent systems that can be applied to job shop floors crowded with small and bigger robots.Now UiT is taking this to a new level were we are addressing self-driving trucks and intelligent baggage handling at airports.  The Covid-19 situation has inhibited a lot of feild work, but now things are being revitalized.  In a new EU initiative the AI and Robotics groups are looking at a new type of service where self-driving trucks can roam the streets in a systematic manner to carry out routine tasks such as sweeping the streets and replace cable-less EV chargers. The work is going to be a cooperation between JOLT, HIAB and Smart Innovation Norway.JOLT and HIAB has already established a cooperation that uses more manual techniques. Now, they are going to ramp it up.

Modern jets are crowded with smart electronics and software that allow a pilot to lean back and let the computer do the flying job from Oslo to New York.  But what happens when the plane is on the ground and baggage is loaded and unloaded?  Well, you have all seen it.  Very little has changed over the past two hundred years.  People are running, crawling and heaving bags and suitcases on and off like they did at docks or railway stations when your great, great grandfather was young. We want to change that.  In the Ali-T Pilot we will initiate work on this.  Already PhD student Andreas has created some cool algorithms and methods for managing robot arms and conveyor belts. 

The work that we know from robotics and AI and machine learning at UiT in Narvik, as well as our common experience from the the job shop floor and from instant response management in the electricity sector will be instrumental. This will be our next contribution as an engineering school to a more intelligent and innovative industry.