Integration and comparison of multi-criteria decision making methods in safe route planner

Motor vehicle crashes are a leading cause of death in the U.S. In order to reduce death and serious injury, road and traffic engineers manually evaluate road segments and visualize the safety level of roads. These existing risk maps can be confusing and must be manually interpreted by drivers to find the safest path from a source to a given destination; this can result in ignoring the safety of the routes by drivers. In addition, common navigation systems such as Google Maps and Waze present two or three alternative paths from a source to a given destination based on the travel time and distance. A navigation system is required to take the safety level of the road segments into consideration while suggesting a path. This navigation system needs to acquire knowledge from various sources, a user interface to obtain user preferences, and an inference engine to find the best paths. Such a system can still suggest multiple conflicting paths, such as shortest, fastest and safest paths. This paper presents the addition of a multi-criteria decision-making (MCDM) method, Analytical Hierarchy Process, to a previously designed Safe Route Planner to aid users in choosing the most suitable path among M alternative paths. Different MCDM methods can generate different results while applied to the same problem. There are a few comparative studies to compare the results of different Multi-Criteria Decision-Making (MCDM) methods. Therefore, a particular attention is devoted to comparing the results of five decision-making techniques, namely AHP, Fuzzy AHP, TOPSIS, Fuzzy TOPSIS and PROMETHEE through two real-world case studies. In addition, the comparative studies fail to adequately quantify the results of the MCDM methods; consequently, another aim of this research is to investigate the applicability of Spearman's rank correlation coefficient, Average Overlap and Discounted Cumulative Gain techniques to quantify the results of the MCDM methods. (C) 2020 Elsevier Ltd. All rights reserved.