Artificial Intelligence (AI) and machine learning influence nearly every facet of daily life. MTRI is a part of the revolution.
Our team of data scientists, physicists, mathematicians, and engineers understand that not every algorithm should be a black box. We believe that machine learning should come with guarantees. It takes much more than inputs and outputs (or transfer characteristics) to solve the hardest problems. MTRI scientists are active in developing state-of-the-art machine learning and statistical techniques for understanding natural and man-made environments.
Projects
Model Inventory of Roadway Elements (MIRE) Road Surface Type Identification
This project's goal was the classification of road types across all of Michigan. Machine learning models were used to classify multi-spectral, multi-resolution imagery of Michigan land to pick out and classify roads into unpaved, asphalt, and concrete. The completed road classification data was given to the Department of Transport.
Railroad Artificial Intelligence Intruder Learning System (RAIILS)
The Railroad Artificial Intelligence Intruder Learning System (RAIILS) project analyzes the effectiveness of AI-based tools for reducing trespass events on rail properties.
Learn more about the initial findings.
Conferences and Workshops
Workshop on Cross-Disciplinary Challenges and Opportunities in AI Applications to Science and Engineering
A collaborative workshop about the common challenges and new frontiers combining science with computing and AI.
Learn more.