Robotics Practical 2

Uitgebreide vaknaam	Robotics Practical 2
Leerdoelen	At the end of the course, the student is able to: 1. Design and implement a supervised learning method (i.e., support vector machine) for the classification of features extracted by a lidar in a simulated environment and in a real lab environment. 2. Design and implement a follow-me control for the navigation of a unicycle mobile robot that follows human’s legs in a simulated environment and in a real lab environment. 3. Design and implement a centralized autonomous coordination control for a swarm of unicycle mobile robots to keep a predefined formation in a simulated environment and in a real lab environment. 4. Design and implement an autonomous control for a unicycle mobile robots for obstacle avoidance in a simulated environment and in a real lab environment. 5. Design and implement an overall system behavior in which the unicycle mobile robots follow human’s legs while keeping a predefined formation and avoiding static obstacles in a simulated environment and in a real lab environment.
Omschrijving	The students gain basic knowledge on the control of a swarm of unicycle mobile robots (mobile base and lidar) to follow a human, while keeping a formation and avoiding obstacles in a simulated environment, and gain practical experience in a real lab environment. Students focus on: - A supervised learning method for the classification of features extracted by the lidar. Students learn: 1) To process simulated lidar data (collect data, cluster them, extract geometrical features); 2) To design and implement a support vector machine on the extracted features to classify human’s legs. Tools: Stage, RViz. - A follow-me control algorithm for the navigation of the robots to follow the human’s legs. Students learn: 1) To compute the central point (on the ground) between two clusters classified as human legs; 2) To design/implement a velocity control for the robots to follow the computed central point by keeping a requested distance. - An autonomous coordination control algorithm for the swarm to keep a predefined formation. Students learn: 1) To compute the centroid of the formation; 2) To design/implement a velocity control for the swarm to keep a requested distance from the centroid. - An autonomous control algorithm for the robots to avoid obstacles. Students learn: 1) To detect additional static obstacles with the lidar; 2) To design/implement a velocity control for each robot to avoid obstacles by keeping a requested distance from them. - An overall system behavior of the swarm. Students learn: 1) To build a control architecture, based on a null-space approach, that assigns priorities to the three tasks (obstacle avoidance, follow-me, formation control); 2) To develop well-structured software. Tools: Python, NumPy, Git.
Uren per week
Onderwijsvorm	Hoorcollege (LC), Opdracht (ASM), Practisch werk (PRC)
Toetsvorm	Opdracht (AST), Verslag (R) (The final weighted grade is based on the practical grade (70%) and the report grade (30%), and should be at least 5.5. In addition, both practical grade and report grade should be above or equal to 5.0. Furthermore, this course has mandatory presence: At least 75% of the lectures/tutorials/practical work is mandatory. This implies that missing more than 25% results in failure of the course.)
Vaksoort	bachelor
Coördinator	Prof. Dr. R. Carloni
Docent(en)	Prof. Dr. R. Carloni
Verplichte literatuur	Titel Auteur ISBN Prijs Lecture notes, tutorial instructions, few research papers.
Entreevoorwaarden	Mandatory: Calculus for AI (WBAI048-05), Linear Algebra and Multivariable Calculus (WBAI050-05). Recommended: Robotics Practical 1, Autonomous Systems, Signals and Systems, Neural Networks, Statistics, Imperative Programming. Prior knowledge of Python and Git is useful, but not required. If the mandatory requirements are not met, only the Board of Examiners of the AI BSc may grant an exemption.
Opmerkingen	A (recent) computer is required to work from home and run a Virtual Machine. This course unit has a capacity limit. More information about capacity-limit courses can be found here. This course has an intended limit of 40 participants.
Opgenomen in	Opleiding Jaar Periode Type BSc Kunstmatige Intelligentie 3 semester II b keuzegroep