Dynamic Vehicle Routing with Indirect Communication via Unattended Ground Sensors
Document Type
Conference Proceeding
Publication Date
1-9-2017
Abstract
The performance of mobile agents responding to dynamic, prioritized requests falls under the broad category of dynamic vehicle routing problems. In the typical formulation, agents receive service requests instantaneously. However, as increasing numbers of au- tonomous agents are employed across large networks there is a need to characterize the performance of these systems in communication constrained environments, where direct contact with the agent may not be possible. This work defines a dynamic, deterministic vehicle routing problem and compares the performance of the agent using three communication schemes: no communication, direct communication, and indirect communication. The goal is to characterize the responsiveness of an indirect, decentralized communication scheme to dynamic task priorities. The presented results show that, under the simulated conditions and for four maps of differing node volumes, the indirect communication mechanism performs on average 3% better than the case with no communication, and under performs the direct communication case by 6%. Furthermore, indirect communication provides the most benefit to agents within the first 50% of their tour. Future work will draw upon this data to develop a decentralized approach for servicing a network of nodes with a team of UAVs.
Source Publication
AIAA Guidance, Navigation and Control Conference
Recommended Citation
Olsen, C. C., & Kunz, D. L. (2017). Dynamic Vehicle Routing with Indirect Communication via Unattended Ground Sensors. AIAA Guidance, Navigation and Control Conference, #2017-1527. https://doi.org/10.2514/6.2017-1527
Comments
This conference paper is available through subscription or purchase from the publisher, AIAA, using the DOI link below.
Conference Session: Aerospace Robotics Control