It’s back-to-school season and Motion Metrics would like to wish our departing co-op students a successful year at their respective institutions as they return to their studies! This semester’s students were an exceptionally talented bunch, and we will miss their friendly faces at our head office in Vancouver.
Each student was assigned a supervisor and semester-long project that culminated in a final company-wide research presentation. To find out what our co-op students have been up to this semester, read the intern profiles below.
Sharan Sankar – Machine Learning Co-op
Sharan is a Computer Engineering student from the University of Waterloo and the winner of our co-op presentation this semester. Under the guidance of a senior machine learning developer, Sharan developed a more efficient pipeline for multi-object detection and tracking leveraging a recurrent neural network (RNN). This centralized and highly modular pipeline architecture will improve detection performance, streamline development, and allow Motion Metrics to service customer feature requests.
Sharan found great satisfaction in aligning his project goals with company objectives and delivering a body of work that adds direct business value to the company. “I have learned a lot from everyone here,” he explains. “Even as a co-op, I can bring value to the company.” Sharan hopes that his experience at Motion Metrics will translate into future employment opportunities during his next co-op semester.
Carter Fang – Systems Engineering Co-op
Carter is a Mechatronics student at the University of British Columbia who worked alongside a senior systems engineer to design and execute experiments investigating the performance of stereovision sensors. He was eager to learn and practice software development skills that weren’t part of the curriculum for his formal education, and his supervisor was happy to oblige.
Carter had a diverse range of responsibilities in his internship and really got his hands dirty building tools for analysis in modern programming languages. “My understanding of stereovision sensors was deepened by the development of camera configuration and calibration tools,” he explains. “Using Python, I implemented software tools and automated hardware systems that standardize the process of camera configuration. Using Google’s testing framework, I also developed unit tests for upcoming products in C++.”
Fatemeh Taheri Dezaki – Machine Learning Co-op
Fatemeh is a doctoral researcher at the University of British Columbia’s Robotics and Control Laboratory who collaborated with a senior machine learning developer to design and test a real-time video classification algorithm. Inspired by human vision, her algorithm learns the operating cycle of a mining shovel by encoding it into a series of activities and annotating video streams using deep neural networks.
Fatemeh’s project has the potential for many exciting applications in future products. Her algorithm could enable us to understand the operating cycles of other machines at the mine site, provide status updates, or add functionality to our stereo cameras. “I was working towards making mining cameras smarter through the use of the state-of-the-art deep learning methods in video perception,” she explains. “Working at Motion Metrics provided me with opportunities for advancement that I would not have otherwise. The environment is very friendly, and I got to work with some great people who are always willing to help.” Fatemeh tested her algorithm on ShovelMetrics™ and has already achieved a high degree of accuracy.
Jess Montilla – Support Co-op
Jess is a third-year Mining Engineering student from the University of British Columbia. Under the guidance of our support team, he studied the downstream effects of fragmentation in mine sites. Drawing on his mining education, Jess opened his presentation with an overview of the open-pit production cycle.
Throughout his semester at Motion Metrics, Jess studied fragmentation data at three of our major mine site installations. His final project summarized the operational efficiencies of the ShovelMetrics™ systems installed at each of the three mine sites and studied the effects of post-blast mechanisms against different performance indices like crusher efficiency, average shovel dig times, bucket payloads, and missing teeth. “During my co-op, I helped evaluate and troubleshoot any in-situ incidents by remotely accessing installed systems and analyzing data to increase productivity on site,” Jess explains. “I gained valuable experience in an unfamiliar environment by applying the theoretical and technical knowledge I learned through my supervisors, and I approached each new challenge with an open mind.”
Want to join us?
Motion Metrics is a great place for ambitious co-op students to take on meaningful, challenging work alongside experts in their field. Our co-op students come from a range of backgrounds, including Electrical and Computer Engineering, Computer Science, Engineering Physics, Mechatronics, and Mining Engineering. Many of our co-op students return to work for us once they’ve graduated.
If you’re a motivated undergraduate or graduate student with the skills to succeed in a fast-paced, high-tech company, check out our current openings for co-op positions here.
Want to learn more?