Towards a multi-criteria, crowdsource-based route planner for recreational pedestrians in Groningen

Abstract

Walking is the most popular recreational activity in the Netherlands and can be characterized by many different trends, such as the use of technology, thematic walking and crowdsource-based route planning. Although these trends have impact on the walking activities in the Netherlands, they are implemented in route planners separately. By focusing on only one trend, the route planners are dismissing the added value of combining the three trends into one route planner. The goal of this research is to combine the three trends and translate them into a multi-criteria, crowdsource-based route planner for recreational pedestrians in Groningen. With the route planner, users can select a route based on their personal preferences, including themes, the start and ending location of the route and the duration of the route. Secondly, users help to optimize the route planner by sharing their feedback on the recommended walking routes with other users. To create the route planner, the focus of this research is on three work packages. The first work package includes the determination of the key capabilities of the route planner, consisting of business requirements, user requirements and system requirements. The list of key capabilities represents all the capabilities the route planner should possess and are based on interviews with recreational pedestrians in Groningen and the research of academic literature, governmental reports and software instructions. The second work package includes the development of a technical framework on how to built a multi-criteria, crowdsource-based route planner in the Web AppBuilder for ArcGIS (Developer edition) software. In this framework, the Vehicle Routing Problem with Time Windows (VRPTW) algorithm approach is explained as well as the feedback mechanism of the route planner. After the deployment of the route planner, the third work package is introduced. During this work step, a test group performed a try-out assignment that included navigating with the route planner, rating the stops along the route and adding new stops to the route planner. Based on the outcomes of the try-out assignment, the route planner is upgraded; the user scores are adjusted and new stops are added to the route planner. In addition, the testers filled in an evaluation form. The outcomes of the evaluation forms are used to reflect on the performance of the route planner, on the research limitations and on the recommendations for further research. By creating a multi-criteria route planner, users can obtain a walking route that closely fits their personal preferences. Moreover, by adding new stops to the route planner, the scope of the route planner expands and users can obtain more varied walking routes. In overall, the results of the testing prove that the VRPTW and the feedback mechanism increase the planning effectiveness and the recommendation accuracy of the route planner.