Kaymie Shiozawa
MIT SoE | Quest Undergraduate Research and Innovation Scholar
Seeking Critical Information for Autonomous Mining and Construction Machines via Operator Gaze Tracking
2018–2019
Mechanical Engineering
- Artificial Intelligence & Machine Learning
Harry Asada
Automating excavation in mining and construction applications is crucial today, as the supply of skilled operators cannot match market demand. To make control decisions for autonomous excavators, gaze-tracking can be employed by solely extracting key information that skilled operators use in the field during a digging task. A subject will wear a camera that looks into both the eye and the surroundings to collect data. After characterizing the human attention points of a digging task, transition points from digging to picking up can be identified using optical flow. This information will serve to identify appropriate inputs in the control decision process.
I am participating in SuperUROP because I would like to gain advanced research skills within robotics to apply to graduate studies. Last year, I was a UROP student in the D’Arbeloff Lab, and I also took 2.671 (Measurement and Instrumentation), a class that has prepared me to present my research results effectively. I’ m excited by the opportunity to learn more about the software aspects of robotics and training neural networks to improve the robot’s intelligence and ability to mimic human behavior.