Improving Operational Resource Efficiencies through the Application of Model-Based Reinforcement Learning (MBRL)

Reinforcement learning (RL) is the problem of designing an agent that interacts with its environment and adaptively improves its long-term performance. Many complex real-world industrial decision-making problems can be formulated as an RL problem. RL is at the core of artificial intelligence and has the potential of having a huge impact on our economy and society, perhaps more so than any other area of machine learning. Model-based Reinforcement Learning (MBRL) is a promising approach to design sample-efficient agents for problems where the number of interactions with the real-world cannot be very large. The goal of this project is to study the feasibility of the MBRL algorithms to solve industrial problems with the help of the project industrial partner. Specifically, the project aims to help industrial participant’s with to meet targets for carbon footprint reduction by applying MBRL approach to the operational processes of concern. A primary industrial application that the project will work on is optimization of the logistics network by routing trucks efficiently to reduce fuel consumption and CO2 emissions.

Faculty Supervisor:

Amir-massoud Farahmand

Student:

Romina Abachi

Partner:

Linamar

Discipline:

Engineering - computer / electrical

Sector:

Manufacturing

University:

University of Toronto