Mobile Robot Systems

Aims

This course teaches the foundations of autonomous mobile robots, covering topics such as perception, motion control, and planning. It also teaches algorithmic strategies that enable the coordination of systems composed of multiple robots. Through the exercises, the students will learn to use ROS (http://www.ros.org), which is a widely used library in both industry as well as academia. 

Lecture topics

Basics of autonomy and sensor-actuator loops; robot motion, kinematics, and control; perception, localization and mapping; navigation and path planning; multi-robot systems; robot learning.

Pre-recorded material is available here: 

https://www.youtube.com/playlist?list=PLaTKfS3-bDpDyOwrxLcQRGxY9XJw33ANo 

Objectives

By the end of the course students should: 

● understand how to control a mobile robot; 

● understand how a robot perceives its environment; 

● understand how a robot plans actions (navigation paths); 

● know paradigms of coordination in systems of multiple robots; 

● know classical multi-robot problems and their solution methods; 

● Know how to use ROS (Robot Operating System, http://www.ros.org). 

System requirements 

The course practicals involve a robotics simulator that requires certain computing capabilities. To ensure a smooth experience and provide eective support, students will need a system with the following requirements. For those students who cannot meet these requirements, we will lend out systems with all tools pre-installed. 

Hard Requirements (Minimum Needed to Run) 

● OS: Ubuntu 22.04 (64-bit) 

● Setup: Native installation (dual boot accepted; WSL, Docker, Virtualization tools not supported) 

● CPU: 2 GHz dual-core 

● RAM: 8 GB 

● Graphics: OpenGL 3.3+ capable 

● Storage: > 64 GB of free space 

Soft Requirements (Recommended for Better Performance) 

● CPU: Quad-core 

● RAM: 16 GB 

● GPU: Dedicated GPU (NVIDIA/AMD recommended) 

● Storage: Additional space for simulations and software 

Assessment 

The assignments will consist of two elements: (1) experimental work using a robot simulator and (2) theory / understanding. The exercises will require data collection and analysis. The balance between practice and theory will depend on the exercise topic. Each student will

submit a written report. Students will be expected to be able to demonstrate any results reported in their hand-in. 

The final mark for this module will be determined in 4 equal parts by two written reports (one for each assignment) and two in-person exams (one for each assignment). The lecturer will hold an in-person questioning session with each student to discuss their submissions. Submissions are non-anonymous. Each written assignment will compose 25% of the final mark; the remaining 50% of the mark will be determined by the student's performance in the two oral examinations with either the lecturer or a senior assessor. 

Recommended reading 

• Siegwart, R., Nourbakhsh, I.R. and Scaramuzza, D. (2004). Autonomous mobile robots. MIT Press. 

• Thrun, S., Wolfram B. and Dieter F. (2005). Probabilistic robotics. MIT Press.

• Mondada, F. and Mordechai B. (2018) Elements of Robotics. Springer

• Siciliano, B. and Khatib, O. (2016) Springer handbook of robotics. Springer.

• Mesbahi, M. and Egerstedt, M. (2010) Graph theoretic methods in multiagent networks. Princeton University Press.