Computer vision algorithms for coordinated robot movement and hose manipulation for vacuuming debris.
Avideh Zakhor, Professor
Electrical Engineering and Computer Science
Applications for Fall 2025 are closed for this project.
Over the past few years, we have developed autonomy algorithms to enable a tracked robot to vacuum debris ( peanuts, insulation etc) as it traverses over joists. The current approach is not adaptive in that the hose moves in a predefined sequence of movements.
In this project, we will develop an adaptive method in which the robot sensors determine if there are debris left behind and based on the location of the debris, our algorithms will position the robot and the hose accordingly to vacuum the remaining residual debris.
You will develop new algorithms and will test it on the raspbery pi 5 on the robot.
Role: * Come up to speed with the existing code base.
* Invent a new algorithm to deal with residual and implement it on the raspberry pi
* test the algorithm in real life and demonstrate it works efficiently.
Qualifications: * Must have taken CS 180 and/or CS 280.
* Must have experience in programming microcontrollers such as raspberry pi
* Highly desirable to have taken EECS106
* Familiarity with ROS is a plus.
* Preference will be given to candidates who can continue the project over the spring semester.
* Prior robot/drone/hardware experience is a plus
* Prior 3D mechanical design experience is a plus.
Hours: 12 or more hours
Related website: https://www-video.eecs.berkeley.edu/papers/avz/revised-icra-2022.pdf
Related website: https://www-video.eecs.berkeley.edu/papers/avz/revised-icra-2022.pdf