Advanced robotics, estimation theory, deep learning for real-time payload monitoring - BC-347
Preferred Disciplines: Master or PhD in Electric, Mechanical or Computer Engineering
Project length: 4 yo 8 months
Approx. start date: As soon as possible
Location: Vancouver, BC
No. of Positions: 1
Preferences: Open to all Canadian universities. Language: english
Motion Metrics International is a high-tech company that provides advanced monitoring solutions to the mining industry. We are currently hiring engineers to scale up our AI infrastructure. Our products apply machine vision and other sensor technologies to provide critical alarms and decisive information to all tiers of users in the mine’s hierarchy.
Shovel productivity can vary dramatically between different operators and loading practices. When a truck is overloaded, component life is significantly decreased which can lead to premature tire wear and suspension failures. When a truck is underloaded, more trips and passes are needed to transport the same amount of material. Payload Monitoring system relies on the position of the shovel bucket and the associated pressures in the hydraulic cylinders to estimate the load inside the bucket. The goal of this project is to research proprietary algorithms and use concepts from advanced robotics and estimation theory, to determine payload in real-time without interrupting the normal flow of operation.
Background and required skills
- Explore proprietary algorithms to automatically determine payload while meeting defined system requirements and constraints
- Define validation and performance metrics for proposed algorithms
- Create documentation, including Internal reports, conference/journal papers
- Combination of Advanced robotics, estimation theory, and deep learning
Expertise and Skills Needed:
- One year of hands-on experience developing and implementing machine learning algorithms
- Understanding of theoretical concepts of statistics, probability, advanced robotics and estimation theory
- Knowledge in signal and image processing techniques
- Proficiency in a modern programming language like Python
- Proficiency in programming language such as C++ or C#
- Experience with tools such as Theano and Caffe
- Proficiency with version control (git, SVN).
For more info or to apply to this applied research position, please