Asef, Pedram; Alexandre, Benoit; (2024) Navigating Intelligence: A Survey of Google OR-Tools and Machine Learning for Global Path Planning in Autonomous Vehicles. Advanced Intelligent Systems , 6 (9) , Article 2300840. 10.1002/aisy.202300840.

Preview

Text Asef_Advanced Intelligent Systems - 2024 - Benoit - Navigating Intelligence A Survey of Google OR‐Tools and Machine Learning.pdf Download (16MB) | Preview

Abstract

This study offers an in-depth examination of Global Path Planning (GPP) for unmanned ground vehicles (UGV), focusing on an autonomous mining sampling robot named ROMIE, which plays a crucial role in geochemical mining sampling. GPP is essential for ROMIE’s optimal performance, as it involves solving the Traveling Salesman Problem (TSP), a complex graph theory challenge that is crucial for determining the most effective route to cover all sampling locations in a mining field. This problem is central to enhancing ROMIE’s operational effciency and competitiveness against human labor by optimizing cost and time. The primary aim of this research is to advance GPP by developing, evaluating, and improving a cost-effcient software solution and web application. We delve into an extensive comparison and analysis of various Google OR-Tools optimization algorithms, designed to address different TSP scenarios. Our study is driven by the goal to not only apply but also test the limits of OR-Tools’ capabilities by integrating fundamental Reinforcement Learning techniques like Q-Learning and Double Q-Learning into our approach. This enables us to compare these basic methods with OR-Tools, assessing their computational effectiveness and real-world application effciency. Our comparative analysis seeks to provide insights into the effectiveness and practical application of each technique, informing future advancements in GPP software. Our findings indicate that Q-Learning stands out as the optimal strategy, demonstrating superior efficiency by deviating only 1.2% on average from the optimal solutions across our datasets. In conclusion, the research shows that QLearning-based algorithms are the most effective, suggesting their significant potential in delivering cost-effcient and robust solutions in real-world mining operations, thereby enhancing the capabilities of autonomous robot like ROMIE.

Download activity - last month

Export as CSVJSONXML

Download activity - last 12 months

Export as XMLJSONCSV

Downloads by country - last 12 months

Export as JSONXMLCSV

Archive Staff Only

View Item