Student projects
Available
No-swap regret in repeated games with unknown constraints
Motivation:Several real-world multi-agent systems such as electricity markets, traffic routing, and multi-robot applications can be described as repeated games with unknown constraints. In electricity markets, there are capacity constraints on the load of each grid line. In multi-robot applications in unexplored environments, the robots have to accomplish a task while avoiding collision with obstacles and (…)
Automating Reward Function Design Using LLMs for Robotic Disassembly of EV Batteries
Background and MotivationThe rapid growth of electric vehicles (EVs) is leading to a surge in the number of end-of-life (EOL) batteries, which require safe and efficient recycling processes. Disassembling battery packs is a complex and hazardous task, where robots can offer advantages in terms of precision, safety, and repeatability. However, existing robotic systems face challenges (…)
Stability Guided Reinforcement Learning for Autonomous Robotic Construction
MotivationWhile automation has been widely adopted in many industries, the construction sector has been relatively left aside. The unpredictable construction site requires frequent replanning of robots at the last minute, which is not the case in an assembly line. Employing robots in construction requires developing new algorithms to enhance their autonomy and intelligence.This project inscribes (…)
Ongoing
Multi-robot Trajectory Planning Under Traffic Interactions
1. Overview:This project focuses on developing trajectory planning solutions for autonomous systems in interactive environments. The goal is to create algorithms enabling each robot to safely reach its destination while navigating complex scenarios, such as multiple autonomous cars at non-signalized intersections or multi-robot systems requiring formation control and coordination. Key challenges include collision avoidance with (…)
Model-free risk-sensitive inverse reinforcement learning
Description:The primary goal of this project is to develop a model-free algorithm for risk-sensitive inverse reinforcement learning (IRL), where the expert’s behavior is driven by optimizing a risk measure rather than the expected discounted reward [1]. While the approach proposed in [1] requires access to the transition law, our objective is to create an algorithm (…)
Previously done
Reward learning from human feedback for the Construction of Spanning Structures
Abstract: This semester project addresses the challenge of building a structure connecting two supports across a gap, using reinforcement learning from human preferences. This approach involves learning a reward predictor from human feedback between pairs of demonstrations of the construction task. After presenting the algorithm used to train the agent with human feedback, the report (…)
Coordination and Control of a Group of Ground Robots
Multi-robot systems are becoming increasingly popular and are being used in more and more applications such as warehouses, agriculture, and transportation. The objective of this project is to control and coordinate a group of ground robots. Controlling a single ground robot requires planning the desired path and computing actuation inputs to maneuver the robot based (…)
Optimal Task Assignment and Collision Avoidance for Mobile Robots
Multi-robot systems provide great benefits in applications such as coordinated search and rescue, large scale agriculture, and efficient transportation of people and goods. When a group of mobile robots operate in shared space coordination between them is crucial. In particular, strategies for task assignment and collision avoidance are needed. Task assignment is the problem of (…)
Game theory for energy communities
MotivationIn order to promote an increase in the use of renewable energies, the European Union is thinking about promoting decentralized approaches, where the citizens are actively participating in the production and storage of energy. Declining prices for solar panels and the ability to sell and buy electricity amongconsumers (it is already possible in some countries (…)
Robot Motion Planning via Mixed-Integer Programming
Motion planning plays a key role in the autonomous control of mobile robots. With the advancement of technology and the introduction of increasingly sophisticated autonomous vehicles, such as self-driving cars, delivery drones, and warehouse robots, complex motion planning objectives and constraints need to be considered. The objective for an autonomous car could be the minimisation (…)
Safe reinforcement learning
Reinforcement learning addresses finding a policy for a Markov decision process (MDP) to optimise a cumulative reward function based on the observations of the rewards and the evolution of the MDP. The application of reinforcement learning to safety-critical systems such as autonomous driving and robotics requires safety, that is, satisfying constraints while learning (e.g. collision (…)
Augmented Robot Reality: Projecting the Virtual into the Real World
Multi-robot systems provide great benefits in applications such as coordinated search and rescue, large scale agriculture, and efficient transportation of people and goods. This motivates research on novel coordination, planning, and control algorithms. To develop such algorithms it is crucial to test ideas rapidly in a setting that is both repeatable and adaptable to different (…)