John Folkesson

Are you a student at KTH interested in doing research in our AUV group? Read this: Master Student Projects,more about SMaRC: SMaRC.   We welcome Master students (from KTH) and try to find you interesting projects for DD2414 or DD2411 as well as thesis projects. 

PhD applications: in general all open positions at KTH are advertised and applied for here: KTH Dotoral Positions, mailing me directly is rather futile as I can only hire through that portal.

A list of my underwater publications is https://www.csc.kth.se/~johnf/AUVPubs.pdf.  The RPL Division's Master thesis project proposals are here:  RPL Master Theses (more/different projects are available just ask me)

Courses:

Probabilistic Graphical Models (DD2420), course responsible, teacher, examiner | Kurswebb

Applied Estimation (EL2320/EL3320), course responsible, teacher, examiner | Kurswebb

The Sustainable Systems and Control Engineer (EL2220), teacher | Kurswebb

Popular Science Research Profile: 

John Folkesson works in the field of mobile robotics, studying how robots can move, perceive their surroundings, and make sense of the world. Over the years, his research has covered many kinds of robots: underwater vehicles, flying drones, outdoor ground robots, autonomous cars, and indoor service robots. Today, the main focus of his group is underwater robotics, a domain where some of the hardest problems in robotics are still unsolved.

One of the biggest challenges underwater is navigation—knowing where you are while building a map of an unknown environment. This problem, known as simultaneous localization and mapping (SLAM), becomes especially difficult beneath the surface. Underwater perception is diffecult due to the limited visibility and lack of landmark features. As a result, recognizing places is often ambiguous and requires careful reasoning.

Sonar, the main sensing tool underwater, behaves very differently from cameras or lidar.  Proper modeling of sonar is an important part of underwater SLAM.  The robot can travel for long distances without any sensor contact to the seabed at all. In these situations, estimating motion becomes highly uncertain, calling for new ideas and new methods. At the same time, future autonomous underwater vehicles will need to operate far beyond the reach of human communication. Missions under polar ice or in the deep ocean may last days, weeks or months without contact. For these robots to be truly useful, they must navigate on their own and accurately reference the data they collect to the true positions.