A fuzzy integer-nonlinear programming approach for creating a flexible just-in-time location-aware service in a mobile environment

Abstract

This study provides a timely and flexible location-aware service (LAS) to mobile users in a dynamic environment. Few previous studies have examined similar concepts. In the proposed methodology, the inaccuracy of user positioning and the uncertainty of manual service preparation are considered and modeled by using fuzzy numbers. These numbers are used as inputs to establish a fuzzy integer-nonlinear programming model applied in identifying the most suitable service location and path. A required service is prepared immediately before a user reaches the recommended service location. To manage simultaneous requests from multiple users, the concepts of fuzzy modeling, route planning, and parallel machine scheduling are combined. Thus, the proposed LAS can distribute multiple users among service locations, thereby enabling users to avoid unnecessary waiting, which is a major problem associated with existing LASs. To assess the effectiveness of the proposed methodology, two experiments were conducted in small areas in Taichung City and Taipei City, Taiwan. The experimental results revealed that the waiting times of users were substantially reduced, increasing the average satisfaction level. However, improving the accuracy of user positioning does not necessarily facilitate achieving a high average satisfaction level.