Research Assistant 2

McGill University

1d • $31 • Onsite

About The Position

The Department of Mechanical Engineering seeks a full-time Research Assistant 2 to support a short-term project focused on Vision-Based Localization and Mapping for Autonomous Navigation in Mill Yard. Log-loading machines operating in mill yards are used to transport logs from mill yard piles to the infeed deck, to unload incoming log transport trucks and to manage the log inventory in the yard. It is expected that coarse navigation in the mill yard will be feasible with GPS: GPS-based navigation will enable localization accuracies within a few meters. To carry out these tasks autonomously, these machines need to have capabilities to localize and map their operational environment, in order to gain situational awareness. We have been working on vision-based semantic SLAM algorithms to imbue the log-loading machines with these capabilities. The RA will continue and expand the work on this topic to further improve the semantic SLAM algorithms, to improve their performance and robustness and to allow their computation on edge devices. In addition, we will explore the possibility of using LIDAR sensing to strengthen the semantic SLAM pipeline. The developed algorithms will be integrated with the perception pipeline developed by our industrial partner FPInnovations and tested in the field.

Requirements

• Master's degree in robotics, specifically semantic SLAM for mill-yard environments.
• Proficiency with ROS2, IsaacSim, Python
• Knowledge of ML, DL and common vision-related libraries

Responsibilities

• Research potential improvements to the current semantic SLAM pipeline.
• Generalize the implementation of the pipeline to make it suitable for edge computation.
• Expand the capabilities of the semantic SLAM to allow for more diversity of objects in the mill yard.
• Validate the performance of the pipeline in real operating conditions
• Adhere to all lab safety practices and department policies.

What This Job Offers

• Job Type: Full-time
• Career Level: Entry Level