TMHP33 Autonomous Robots, 3 ECTS-points
/Sensorstyrda autonoma robotar/

Advancement level:
C

Aim:
Extensive international research work is in progress regarding conditions for different kinds of "robots"(in a very wide sense) in aspects of natural science and information technology. The aim of the course is to choose such parts of this research as can be presented to students of technology. The models and methods of the course are general - how to extract information about the movements and shape of an object from sensor signals.

Prerequisites:
Basic courses in mathematical statistics, automatic control and engineering mechanics are compulsory. The course in modelling and simulation should be studied in parallel.

Course organization:
Teachin is in the form of lectures, classes and laboratory exercises.

Course content:
(to be chosen form the list below) Summary of different kinds of robots, craft for offshore, dangerous environments, space and different experimental robots. The concept of autonomy and some examples of system functions/part tasks. Motion models for robots. Number of degrees of freedom. Rate gyros and accelerometers as absolute sensors, principles of inertial navigation. Rate gyro recoupling. Estimation of collision point and collision signatures from accelerometers in prehensive tools. Models of signals from sensors measuring geometric magnitudes; Pulsed laser, Doppler laser, "distant camera", computer vision, coherent ultra-sound, radar , collision sensors. Stochastic geometry for modelling of form of object. Not a course in sensor technology but in sensor modelling. Estimation of motion from non-contact sensors (by use of Kalman filters among other things). The qualities of estimates as a function of physical parameters. The relations between shape of object and accuracy of the motion estimate. Principles for feedback of sensors in gripping/docking and in curve/contour following. Beacon-navigation: "Luleå Turbo Turtle". Modelling of geometric a priori information and description of the task of the robot: Numeric/symbolic information. Different types of adaptivity. "Intelligent" surveillance systems and signal interpretation when multiple sensors are used. Some recent research projects.

Course literature:
Commented articles from journals, compendium.

UPG1Compulsory tasks to be handed in, oral presentation , 2 p.
Course language is Swedish.