Chapter 10 - Spatiotemporal Big Data-Driven Vessel Traffic Risk Estimation for Promoting Maritime Healthcare: Lessons Learnt from Another Domain than Healthcare

Abstract

With the rapid development of maritime industries, the vessel traffic density has been gradually increased leading to increasing the potential risk of ship collision accidents in crowded inland waterways. It will bring negative effects on human life safety and global maritime economy. Therefore it is of vital significance to study the risk of ship collision using big data mining techniques. The big data–driven computational results are beneficial for guaranteeing smart maritime healthcare in the fields of ocean engineering and maritime management. This chapter proposes to quantitatively estimate the ship collision risk based on ship domain modeling and real-time vessel trajectory data. In particular, the trajectory data quality is first improved using the cubic spline interpolation method. We assume that the ship collision risk is highly related to the cross areas of ship domains between different ships, which are then computed using the Monte Carlo simulation strategy. For the sake of better understanding, the kernel density estimation method is finally adopted to visually generate the ship collision risk in maps. Experimental results on realistic spatiotemporal big data have illustrated the effectiveness of the proposed method in crowded inland waterways.