The promises and perils of Automatic Identification System data

Automatic Identification System (AIS) is used to identify vessels in maritime navigation. Currently, it is used for various commercial purposes. However, the abundance and lack of quality of AIS data make it difficult to capitalize on its value. Therefore, an understanding of both the limitations of AIS data and the opportunities is important to maximize its value, but these have not been clearly stated in the existing literature. This study aims to help researchers and practitioners understand AIS data by identifying both the promises and perils of AIS data. We identify the different applications and limitations of AIS data in the literature and build upon them in a sequential mixed-design study. We first identify the promises and perils that exist in the literature. We then analyze AIS data from the port of Amsterdam quantitatively to detect noise and to find the perils researchers and practitioners could encounter. Our results incorporate quantitative findings with qualitative insights obtained from interviewing domain experts. This study extends the literature by considering multiple limitations of AIS data across different domains at the same time. Our results show that the amount of noise in AIS data depends on factors such as the equipment used, external factors, humans, dense traffic etc. The contribution that our paper makes is in combining and making a comprehensive list of both the promises and perils of AIS data. Consequently, this study helps researchers and practitioners to (i) identify the sources of noise, (ii) to reduce the noise in AIS data and (iii) use it for the benefits of their research or the optimization of their operations.