John Folkesson

goofy

 
 

I am an Assistant Professor and Docent at  the Centre for Autonomous Systems at the Royal Institute of Technology, and the Department of Computer Science and Numerical Analysis. I was formerly a research scientist at the Massachusetts Institute of Technology

My research is concentrated on  data fusion and world modeling for mobile robotics,  particularly for navigation.   The idea is to make machines that can learn how to find their way around without the need for guidance. 

To do this the robot needs to form models of the environment suitable to the task.  That might be simply avoiding obstacles or, more complicated tasks like finding its way back to a location previously visited.   These models need to be learned and refined based on the observations made by the robot.  It is this learning that I am focused on

Current Projects:

EU Integrated Project STRANDS: I am one of the participants of the consortium of 6 universities and two industrial partners. We are looking at the problem of long term learning by robots in real world settings.

VINNOVA Project IQMatic: I am a participant along with Linköping University, Scania, SAAB, and Autoliv. The goal is autonomous driving of tracktor trailer trucks.

Current Teaching:

EL2320/EL3320 Applied Estimation, (Master/PhD. Level) 7.5 hp

DD1334 Database Technology (Undergraduate) 6.0 hp .

Current PhD Students: 

Nils Bore, Main Supervisor
Eric Ward, Main Supervisor
Johan Ekekrantz, Co-Supervisor
Akshaya Thippur Sridatta, Co-Supervisor
Rares Ambrus, Co-Supervisor
Rasmus Göransson, Co-Supervisor

Current Master Students: 

Jeremy Pouech Fredrik Hillnertz
Viktoria Alexeeva
Mikael Hedberg
Niklas Lundborg
Vivetha Natterjee
Talani Bachir

Past PhD Students: 

Alper Aydemir, "Exploiting structure in man-made environments", main supervisor Patric Jensfelt,
Jacques Leedekerken, "Mapping of Complex Marine Environments using an Autonomous Surface Craft", (at MIT), supervisor John Leonard

Past (or soon to be) Master Students: 

Joakim Bergius, "Incremental road map generation in three dimensions "
Shuang Xiao, "Learning Motion Patterns in People Tracking"
Lukas Gratwohl, "Improving Autonomous Navigation by Exploiting Dynamic Elements in an Indoor Office Environment"
Daniel Bug, "Occulus Rift Control of a Mobile Robot"
Daniel Goucher, "Database optimization"
Hugo Blanc, "Inertial and optical hybrid head tracking for helicopter simulators"
Bashar Mengana, "Path Planning for Autonomous Heavy Duty Vehicles in Unknown Environments, using Graph Search"
Karl Hansson, "Data-Driven Analysis of the Fuel Saving Potential of Road Vehicle Platooning"
Jesper Köhl, "Positioning Algorithm and Positioning Server for a Positioning System"
Affan Pervez, "Movement Skill Acquisition using Imitation and Reinforcement Learning"
Favio Saponara Iriarte Paniagua, "Object Recognition using the Kinect Object Recognition using the Kinect"
Rasoul Mojtahedzadeh, "Robot Obstacle Avoidance using the Kinect"
Hunter McCelland, "Using Negative Information in SLAM", (at MIT)

My   CV and   publications

My background includes graduate studies in Physics at the State University of New York at Stonybrook,  some work as a researcher at an electronics company in San Jose CA,  and  my own consulting business were I did statistical data analysis.

I have a Bachelor Degree in physics from the City University of New York.   And a  Masters in Computer Science from the Royal Institute of Technology in Sweden .  My  masters thesis  was entitiled Projection of a  Markov Process with Neural Networks.

I defended  Doctorial thesis on October 4, 2005.  Here is what the final version looks like  PhD thesis .

I have designed and implemented  a system for our outdoor robot, Pluto.  You can read about that here  pluto.pdf .    Pluto can  localize to a map to and goto a designated coordinate avoiding obstacles along the way.  Pluto can automatically explore and map an area using a laser range scanner to located the walls of buildings.   The robot can also investigate a parked car: approach it, extend its arm to position a camera to see under the car and then circle the car holding the camera always at the proper angle. 

Others in the lab have integrated with my system to add road  following, person following and simple arm movements.   A number of nice user interfaces have also been developed.  The work goes on but I am no longer directly involved. 

I also helped develope CURE (CAS Unified Robot Environment).  CURE  is a collection of libraries in C++ to help people working in the lab. 

I have also worked on underwater SLAM.    A relatively simple robot must find its way in a confusing underwater world.  Its a task that requires careful modeling of all uncertianties.  Only then can the robot carry out its mission.  

Also:  I play piano and have a patent for a music keyboard related invention.

 
 

 

Contact details
John Folkesson.
CAS/CVAP KTH
Teknikringen 14
Stockholm, Sweden
 

Phone 08-790-6201

Fax None

Email johnf@csc.kth.se
Movies

Papers


dumbo

coyote

pluto
coyote



Generated by John Folkesson, johnf@csc.kth.se Dec. 18, 2010