Integrated ground vehicle and drone routing with simultaneous surveillance coverage for evading intentional disruption

• Published in: Transportation research. Part E, Logistics and transportation review Logistics and transportation review, 2023-10, Vol.178, p.103266, Article 103266
• Main Authors: Zandieh, Fatemeh, Farid Ghannadpour, Seyed, Mahdavi Mazdeh, Mohammad
• Format: Article
• Language: English
• Subjects: ALNS; Cash-in-transit; Intentional disruption; Risk management; Surveillance drone; Vehicle routing problem
• ISSN: 1366-5545; 1878-5794
• DOI: 10.1016/j.tre.2023.103266

[Display omitted] •A new dynamic integrated ground vehicle and drone routing with simultaneous surveillance coverage.•Considering a comprehensive coordination between ground and aerial vehicles.•Utilizing drone surveillance to assess the risk posed by an intentional disruption of the transportation system.•A new overlapped time windows for the simultaneous visit of a drone via two or more ground links.•Developing a new developed ALNS/SA based hybrid algorithm. The purpose of this study is to examine the use of drones to provide transportation security during times of intentional disruption. Due to the nature of valuable goods, their transportation is always associated with certain risks, necessitating implementing security measures to ensure the transportation process’s security. This study aimed to address the security of valuable goods transportation by proposing a new routing problem using drones termed integrated routing and surveillance (IRS) for ground and aerial vehicles operating in dynamic environments. This approach addresses the problem of deliberate disruption in the planning and implementation phases. Risks are estimated during the planning phase using historical data, and routing is designed to minimize risk and travel costs. Drones conduct online surveillance of the solution routes generated during the planning phase throughout the implementation phase. Ground vehicles are only permitted to enter a link if drones assess the risks before their entry. Drones’ arrival times for this assessment of each ground link are determined by non-predefined time windows. Drones can visit multiple ground links if their time windows overlap in the proposed problem. If the drones detect a suspicious agent, they will perform dynamic replanning to avoid the risk source. Additionally, the proposed problem is optimized using new ALNS (adaptive large neighborhood search) and SA-based (simulated annealing) algorithms. The proposed algorithm’s efficiency and effectiveness are evaluated for small and large cases. The results demonstrate the proposed model’s applicability and superiority to ALNS.