A data fusion approach to predict shipping efficiency for bulk carriers

Abstract

Maritime waterways are critical transportation systems that connect economies and manufacturing centers. Growing demand for freight movement, along with industry commitment to minimize its environmental impact, has increased emphasis on port and vessel efficiency. Yet, few objective performance measures exist to inform decision making for system improvements. There is an existing gap in quantifiable and objective metrics for maritime transport systems which motivated this work to investigate waterway performance efficiencies through big data analytics. Availability of big data affords practitioners and researchers the opportunity to develop new performance-based metrics to improve maritime logistics. This study focused on short sea shipping logistics of iron ore in the Great Lakes and makes three fundamental contributions. Principally, we propose a maritime transport efficiency (MTE) metric attained through fusion of data from the Automatic Identification System (AIS) and navigation lock data that integrates travel time and vessel payload. We present a linear model to predict vessel capacity based on water surface elevation which will enable practitioners to better adapt to seasonal changes and dredging needs specific to the Great Lakes. Additionally, we present travel time statistics for bulk carriers on the waterway observed through historical AIS data which extends the body of knowledge from earlier works and establishes a reference for system performance. Techniques presented here are effective in capturing travel time statistics in a non-linear interconnected system. This data-driven approach offers new insights for logistics planning and optimization with direct applications to short sea shipping and inland waterways systems. These insights to port and fleet performance allow for querying and simulation of cost impact from investment strategies aimed to improve efficiency or maximize value for operational expenses.