IHA-4306 Fundamentals of Mobile Robots, 5 cr
Additional information
Opintojaksolla IHA-4306 Fundamentals of Mobile Robots rajoitamme opiskelijoiden määrää
The number of students will be limited.
The course is only intended for degree students
Person responsible
Reza Ghabcheloo, Risto Ritala
Lessons
| Implementation | Period | Person responsible | Requirements |
| IHA-4306 2018-01 | 1 - 2 |
Reza Ghabcheloo Risto Ritala |
Individual and group assignments passed. Open book exam. |
Learning Outcomes
This course introduces some answers to the basic questions of "where am I?" and "where have I seen?" and "How do I get there?", which are called location, mapping and planning, respectively, in the robotic community. More specifically, - Students will learn basics about range sensors (Lidar, Radar, Sonar), radio based (GNSS, UWB), egomotion sensors (IMUs, wheel odometry) and their noise characteristics and probabilistic modeling. - Students will learn about coordinate frames and sensor kinematics, that is, how to calculate sensor output in different coordinate frames. - Students will learn how to fuse information comming from different sources (sensors, maps, control inputs,etc) using Bayes filters in particular Kalman filters and particle filters, and to use those to localize moving platforms. - Students will learn about basic world model representations and how to build them (map building) from sensor inputs. - Students will learn important deterministic route planning methods: Dynamic programming (DP), Dijkstra, A*, - Student will also learn planning under uncertainty with MDP (Markov Decision Processes) Notes: * The focus of the planning part of the course will be on point robots, to avoid some complications which will rise due to differential kinematics of contact with ground. * Although the focus of examples and presentation is on mobile robots moving on a 2D plane, most of the methods are applicable to higher dimenstions (manipulators or moving in 3D).
Content
| Content | Core content | Complementary knowledge | Specialist knowledge |
| 1. | Deterministic planning methods: dynamic programming, Dijkstra, A* | planning under uncertainty: solving MDP | How to convert continuous real world to discrete world suitable for planners. POMDP |
| 2. | Bayesian filtering | mathematics of probabilities | |
| 3. | Localization using Kalman filters and particle filters. | Simultaneous Localization and Mapping | |
| 4. | Occupancy grid mapping | world models and representations | |
| 5. | Sensor technologies (LiDAR, Radar, Sonar, Leddar, IMU, GNSS, UWB), sensor models (LiDAR, Sonar, IMU) and their uncertainty | ||
| 6. | Motion control: path smoothing, path following and trajectory tracking | state space models, state feedback |
Additional information about prerequisites
Good programming skill (Matlab or C/C++), good maths (matrix algebra, probability theory), algorithmic thinking, dynamic systems and feedback control
Correspondence of content
There is no equivalence with any other courses