A distributed IoT-enabled MCDM framework for dynamic route optimization of dangerous goods transportation

Logistical challenges arise with the transportation of dangerous goods. This research addresses this problem taking into consideration the essential balance between cost, duration and risk, to ensure a safe and efficient transportation in sustainable transportation systems. The main contribution is to ensure an optimal route selection for both static environmental factors and the inherent uncertainties of dynamic events captured via Internet of Things (IoT) sensor networks that can necessitate real-time re-routing. Current centralized approaches often struggle to provide such robust route evaluations due to latency and data processing bottlenecks in dynamic environments. Therefore, a distributed Multi-Criteria Decision-Making (MCDM) framework is considered to help in evaluating available transportation routes at the network edge and then decide on the optimal one. Several distinct MCDM methodologies are used to identify variations in these potential paths under two ultimate environmental conditions. The static one where transportation conditions are constant and no changes arise and the dynamic one where the probability of disruptive incidents is incorporated based on real-time streaming data from the IoT nodes. In this work, routes are chosen based on TOPSIS method with weights calculated by those different MCDM methods integrated into a distributed architecture. The outcome is then compared with the selected path when using Dijkstra with the same weights. This work aims to compare the outcomes of the MCDM methods used to predict the best route for the transportation of dangerous goods, offering a robust, scalable and efficient tool for optimizing logistics in this critical domain. Each and every method outcome is compared with the outcome of Dijkstra algorithm.