Comparing the accuracy of estimated time of arrival in mobile navigation applications

Abstract

Accuracy of estimated time of arrival (ETA) for mobile vehicle tracking applications (apps) remains as a key issue to be discussed. Most of the mobile vehicle tracking apps such as WAZE is using GPS to track the current location of the vehicles. However, the accuracy of data provided by the mobile apps and factors affecting the accuracy remain as an unexplored area. This study examines the ETA accuracy by comparing ETA of a self-developed vehicle tracking app namely UCSITravel and a popular navigation app namely WAZE. The comparison test is performed by using t-test to determine if there is a statistically significant difference between the means of the datasets. Factors affecting ETA accuracy are determined in this study. The comparison results show that UCSITravel, which uses mean-value calculation (MAP), produces as accurate ETA as WAZE, which calculates ETA based on historic and real-time crowdsourced traffic data. The results shed light to mobile application developers to explore and understand more options of prediction models in developing mobile vehicle tracking app. In addition, this study has demonstrated several factors that were found to affect the accuracy of ETA in both apps as in line with literature. The results yield theoretical and practical implications.