Robots working in human environments will often encounter tasks like cleaning apartments/o ce oors, cooking food, moving heavy items, etc. To perform such high level tasks, robots must be capable of executing lower level tasks such as collision avoidance, climbing stairs, passing through doors, etc. This thesis attempts to address one of the aforementioned tasks; approaching and opening a door. The door opening task have been extensively researched with mobile manipulators, but these robots lack capabilities to maneuver a human centric environment that is lled with elements like staircases and uneven terrains. Humanoid robots, on the other hand, are designed to have anthropomorphic structure giving them the capability to navigate through such elements and perform tasks with higher dexterity. The door opening task is performed with a miniature humanoid robot, DARwIn-OP. A pre- planned trajectory method is developed to test the capability of the robot to perform the task. A trajectory is planned on simulation and then tested on the hardware platform. A semi-autonomous method is also developed to perform the task. 2D and 3D computer vision algorithms aid the robot to detect and approach the door autonomously. Then, a human operator controls the robot to perform the door opening task. Video feed from the robot's vision system is utilized to carryout the task. The performance of the semi-autonomous method was compared with a complete tele-operated approach and similar success rates were observed.
Metrics
37 File views/ downloads
21 Record Views
Details
Title
Door Opening by a Miniature Humanoid Robot
Creators
Bharadwaj Ramesh - DU
Contributors
Paul Yu Oh (Advisor) - Drexel University (1970-)
Awarding Institution
Drexel University
Degree Awarded
Master of Science (M.S.)
Publisher
Drexel University; Philadelphia, Pennsylvania
Resource Type
Thesis
Language
English
Academic Unit
College of Engineering (1970-2026); Electrical (and Computer) Engineering [Historical]; Drexel University