Commentary | IT Matters

How to Make Ship Arrival Analytics and Predictions More Precise

Tags: Logistics I.T., Ocean, Ports, Electronic Data Interchange (EDI), Global Trade Management

Tyson Leal is Director, Industry Sales and Business Development at TransVoyant, 703-778-3500.

Ninety percent of the world’s logistics traffic moves via ocean transport, and until today, the entire logistics community of shippers, forwarders, carriers, and tracking technology providers has suffered from the struggle to consistently predict the precise arrival of the world’s ocean vessels with reliable certainty. Many in the supply chain industry claim that dynamically predicting the precise arrival of the world’s ocean vessels is too complex, requires non-existent data, and is simply unobtainable. Unfortunately, the “good enough” levels of certainty that have been used in the past result in unnecessary spend for all involved.

Alternatively, the ability to accurately predict the arrival for the world’s ocean vessels by lane, by carrier, by month, by day, by hour or even by minute does exist. The data needed can be collected agnostically and results in predictive analytics that provide certainty levels eclipsing 95 percent.

The demands of buyers and technology providers in the ocean transport spectrum are changing fast and forcing organizations to look beyond EDI to reduce unnecessary spend. Traditional event systems provided by marquee software companies have found themselves unable to meet the demand for real-time updates and high accuracy predictive analytics that impact global trade.

Traditional EDI tools are only as good as the data received from partners. That means these tools are driven by self-reporting processes aimed at best managing key performance indicators (KPIs), not data quality or accuracy measures. Additionally, systems built on the foundation of EDI can be best described as reporting “yesterday’s news” and were not designed to deliver the high velocity and high accuracy predictive analytics needed to address the evolving problem set of certainty in the supply chain.

Solving for More Shipping and Supply Chain Certainty

To combat the uncertainty of the EDI-based network data and carrier performance issues, some organizations have subscribed to services that utilize the Automatic Identification System (AIS). This system was developed and is now required by the International Maritime Organization for all commercial vessels over 300 tons to transmit location data. The problem subscribing companies have found with AIS is that its metadata outputs and visual dots on a map aid very little in understanding the critical components of vessel/cargo arrival.

Requirements for Better Vessel-Related Analytics

Organizations that understand the operational value associated with highly accurate predicted vessel arrival times have identified that to deliver this insight requires a different approach. That approach combines three key elements that fuel better analytics: 

  1. Global “Live” and Predictive Data. To predict the arrivals of the world’s ocean vessels, a solution must have the ability to stream the world’s live and forecast data. This includes key sensor-generated feeds, live weather trends and forecasts, social media, port activity, trade lane congestion, vessel dwell times, sailing schedules, natural disaster updates, news feeds, and many others. 
  2. High-Accuracy and High Speed Analytics. Every moment of the transport process is monitored, amassing terabytes of structured and unstructured data and events. Data inflows are less than half the battle when it comes to supply disruption predictions. All of this data must be normalized, analyzed, and have the proper mathematics applied to provide the insight needed for organizations’ systems of record or execution to take action. Through leveraging high-speed analytics, the next generation of predictive analysis will shorten decision times from days/weeks to minutes/seconds, and continuously raise certainty levels in the process.
  3. Scalable Data Science in Real Time. The last and most critical component of the next generation approach to predictive vessel analytics is the ability to apply machine-learning processes to further improve accuracy. Every monitored event and moment is rich with live and historical trend comparisons that have the power to improve operations. Through crowd, e-commerce, and custom analytics, next level systems will leverage every aspect of the transport process to improve operations worldwide. 

When combined, these three components provide the power to turn predictions into actions that decrease variability in a previously uncontrollable process.