Object Detection and Collaboration in Heterogeneous Multi-Robot Systems

With the development of novel technologies, robots are becoming more widely used in everyday applications. Compared with a single robot, Multi-Robot Systems (MRS) are more powerful and cost effective. This thesis will introduce a complete heterogeneous distributed MRS which contains an autonomous mobile robot (Pioneer 3-DX) and a stationary humanoid robot (Baxter), which are able to jointly perform a task that involves navigation and object manipulation. The contributions of this thesis are: 1) we developed an object detection system, locating the pose of objects with a depthcamera through OpenCV and the HSV color model; 2) we optimized a path planning and control system which would enable the Pioneer 3-DX to map its surrounding environment and navigate to target poses, as well as allow the Baxter to grasp, lift and drop the target objects using the MoveIt framework accurately and successfully; 3) we developed a distributed communication system so as to allow two different functional robots to communicate with each other through ROS 2. These contributions are validated with a heterogeneous robot team, including an autonomous mobile robot and a humanoid robot, completing tasks collaboratively.

Level thesis:degree_level

Master's Degree

Year dc:date.issued

2022

Author dc:creator

• Jiang, Song

Advisor dc:contributor.advisor

• Nicolescu, Monica

Committee members dc:contributor.committeemember

• Nicolescu, Mircea
• Panorska, Anna

• MRS
• Multi-Robot Systems
• Robot Collaboration
• Robotics

Handle dc:identifier.uri

http://hdl.handle.net/11714/8175

OAI identifier oai:identifier

oai:scholarwolf.unr.edu:11714/8175