Off-campus WSU users: To download campus access dissertations, please use the following link to log into our proxy server with your WSU access ID and password, then click the "Off-campus Download" button below.
Non-WSU users: Please talk to your librarian about requesting this dissertation through interlibrary loan.
Access Type
WSU Access
Date of Award
January 2024
Degree Type
Dissertation
Degree Name
Ph.D.
Department
Industrial and Manufacturing Engineering
First Advisor
Saravanan Venkatachalam
Abstract
Industry 4.0 concepts and technologies, such as artificial intelligence (AI), machine learning, and cloud computing, are revolutionizing various fields. These advancements are transforming the application of autonomous vehicle technology in both civilian and military contexts. Unmanned ground vehicles (UGVs) and unmanned aerial vehicles (UAVs), previously controlled manually or remotely, now operate with increasing autonomy, giving rise to autonomous mobile robots (AMRs). AMRs use sophisticated sensors, AI, and machine learning to navigate and understand their environments independently. They are employed in diverse civilian and military applications, including agricultural spraying and harvesting, nuclear plant operations, crowd control, environmental monitoring, disaster management, and intelligence, surveillance, and reconnaissance (ISR) missions such as explosive ordinance disposal (EOD), firefighting, border surveillance, and search and rescue.
The US Army is actively developing advanced military ground vehicle technologies, focusing on autonomous navigation, robotic autonomous mobility, and human-robot interfaces. Initially developed to assist ground forces in transporting heavy equipment, AMR technology has evolved significantly. Today, AMRs collect data, including visible/infrared/thermal images, videos of points of interest (POIs), and environmental data such as temperature, moisture, and humidity, using onboard sensors and delivering them to a base station. They are particularly useful in unsuitable terrains, harsh environments, and tedious information collection processes. Small UGVs, for instance, can deliver supplies to remote or hostile areas, reducing risks to defense personnel and increasing combat effectiveness. These capabilities have the potential to revolutionize Army operations on the battlefield.
Despite these advantages, autonomous vehicles face several challenges. Their limited payload capacity and range necessitate multiple stops for refueling or recharging, as well as security halts or mission handovers. Additionally, they must operate within limited mission timeframes. Changes in terrain, obstacles, asset failures, and sensor malfunctions contribute to uncertainty in vehicle availability. The degree of autonomy varies by mission, ranging from fully remotely operated to fully autonomous. Effective human-robot collaboration is crucial, particularly in semi-autonomous operations where human oversight is required so that human drivers can be reserved for critical tasks.
Path planning for autonomous vehicles involves high-level and low-level planning algorithms. High-level mission planning algorithms determine the sequence of POIs and refueling stations AMRs must visit, considering environmental uncertainties and operational constraints. Given their limited range, autonomy levels, and mission uncertainties, formulating and solving these high-level planning problems efficiently is critical. These problems referred to as vehicle routing problems(VRPs) in literature being combinatorial and NP-hard, are typically solved offline before mission commencement.
The goal of this research is to develop optimization models and computationally efficient methods for high-level mission planning, advancing the state-of-the-art in path planning, optimal control, and combinatorial optimization. This research focuses on implementing the best course of action during the planning and operational stages of deploying teams of AMRs. By addressing the challenges associated with autonomous vehicle deployment, this work aims to maximize their potential benefits and expand their range of applications.
Recommended Citation
Chirala, Venkata Sirimuvva, "Novel Data-Driven Algorithms For Autonomous Vehicle Path Planning During Planning And Evaluation Stages" (2024). Wayne State University Dissertations. 4072.
https://digitalcommons.wayne.edu/oa_dissertations/4072