Quantifying the potential of data-driven mobility support systems

Abstract

When traveling it is often necessary to take a detour, for example to find an on-street parking opportunity or a charging station. Numerous systems intending to reduce time or other resources spent on such detours have been presented. An example are methods guiding drivers to free on-street parking opportunities. However, the question of how much can actually be saved by using such solutions when compared to the status quo remains largely unanswered. Often, the cost attached to these detours is unclear. In this work, we present a generalized approach to answer these questions: A methodology consisting of an evaluation environment powered by real-world data and implementations of different scenarios. We then illustrate our proposal by using it to quantify the potential of an optimal assistant for finding on-street parking opportunities. We further show how to generate synthetic but realistic parking data when real-world data is not available.