Making sense of the physical world with high-resolution tactile sensing
Add to Google Calendar
With the rapid development in robotics, people expect robots to be able to accomplish a wide variety of tasks in the real world, such as working in factories, performing household chores, and caring for elderly adults. However, it is still very difficult for robots to act in the physical world, where one of their major challenges is perception, particularly the lack of adequate tactile sensing. In this talk, I will discuss my work on building an intelligent tactile sensing system to help robots better understand and interact with the physical world. My work addresses this question by using a high-resolution tactile sensor and developing algorithms to interpret its output. My work uses a vision-based tactile sensor (GelSight), which measures the geometry of object it touches, as well as the traction field on the contact surface. My work on interpreting the sensor's output has used both traditional statistical models and deep neural networks.
In order to make tactile sensing useful for robots acting in the world, I study both robotic exploration and manipulation tasks. For exploration, I use active touch to estimate the physical properties of the objects. The work has included learning the basic properties (e.g., hardness), of artificial objects, as well as estimating the general properties of natural objects via autonomous tactile exploration. For manipulation, I study the robot's ability to detect slip or incipient slip with tactile sensing during grasping. My research will help robots to more thoroughly understand and flexibly interact with the physical world, allowing them to become more capable of intelligently completing a wide variety of tasks.
Wenzhen Yuan is a Ph.D. candidate in the Department of Mechanical Engineering at MIT, affiliated with Computer Science and Artificial Intelligence Laboratory (CSAIL). Her dissertation research is supervised by Prof. Edward Adelson and Dr. Mandayam Srinivasan. Her research interests include robotic tactile sensing, as well as robotic perception in other modalities. She received her Master of Science degree from MIT, and Bachelor of Engineering degree from Tsinghua University.