Giving robots a light touch : unobtrusive tactile perception in unstructured spaces
Abstract/Contents
- Abstract
- Advancements in robotics are leading toward their integration into domestic settings for the automation of household chores. However, the inherent complexity of unstructured environments, such as homes, poses a formidable challenge for robots. Unlike controlled industrial settings, homes exhibit diverse layouts, variable lighting conditions, and occlusions that render conventional vision-only perception systems inadequate. Mundane yet crucial tasks, like retrieving an item from the rear of a refrigerator, underscore the limitations of solely relying on visual cues. This thesis introduces tactile sensing strategies to enhance robots' performance in unstructured environments. Specifically I demonstrate three approaches that aim to provide robots with new tactile capabilities: (i) a soft distributed tactile skin for safe navigation in spaces with obstacles, (ii) whisker-inspired sensors for non-intrusive object perception, and (iii) low-inertia grippers for low-impact interactions. First, I investigate how to ensure gentle interaction with objects by introducing a soft tactile skin with distributed sensors. This tactile skin helps robots to safely navigate tight spaces by reducing contact forces. By giving robots the ability to detect light contacts and move so as to keep those contact forces small, the skin can mitigate the effects of frequent unplanned contacts (e.g. tipping objects) in cluttered environments. Second, I investigate how robots can leverage contacts to sense their environment. I introduce whisker-inspired sensors that allow robots to sense objects through continuous gentle contact. Non-intrusive contact with objects enables robots to perceive distant objects and partially determine their shapes. With repeated exploration of the environment, the tracked contacts allow the robot to build an understanding of the environment that is robust to visual occlusions. Last, I investigate how low-impact contacts can be leveraged to enable robots to explore and manipulate objects in the environment. I introduce a low-inertia gripper design that enables robots to confidently locate objects by contacting them, determine some of their properties (e.g., weight), and grasp them. In summary, this thesis explores ways to use soft and light tactile sensors and end-effectors to increase the ability of robots to perceive and operate in the cluttered and contact-rich environments.
Description
| Type of resource | text |
|---|---|
| Form | electronic resource; remote; computer; online resource |
| Extent | 1 online resource. |
| Place | California |
| Place | [Stanford, California] |
| Publisher | [Stanford University] |
| Copyright date | 2023; ©2023 |
| Publication date | 2023; 2023 |
| Issuance | monographic |
| Language | English |
Creators/Contributors
| Author | Lin, Michael Andres |
|---|---|
| Degree supervisor | Cutkosky, Mark |
| Thesis advisor | Cutkosky, Mark |
| Thesis advisor | Bohg, Jeannette |
| Thesis advisor | Kennedy, Monroe |
| Degree committee member | Bohg, Jeannette |
| Degree committee member | Kennedy, Monroe |
| Associated with | Stanford University, School of Engineering |
| Associated with | Stanford University, Department of Mechanical Engineering |
Subjects
| Genre | Theses |
|---|---|
| Genre | Text |
Bibliographic information
| Statement of responsibility | Michael Lin. |
|---|---|
| Note | Submitted to the Department of Mechanical Engineering. |
| Thesis | Thesis Ph.D. Stanford University 2023. |
| Location | https://purl.stanford.edu/yv446vn9633 |
Access conditions
- Copyright
- © 2023 by Michael Andres Lin
- License
- This work is licensed under a Creative Commons Attribution Non Commercial 3.0 Unported license (CC BY-NC).
Also listed in
Loading usage metrics...