Puzzling Through Predictive Analytics

To Make the Most of Your Data, Start by Turning Over the Pieces
ink drop

Imagine putting together a puzzle – with the pieces all upside down. You have a lot of pieces, but no clue how to put them together. That is what happens when you have a lot of data, but not insight on how to make it useful. When it comes to data, it isn’t just about having a lot of it. It is about how to share it, protect it and use it to make informed decisions – finding out how each piece can fill a spot and collaborate with the pieces around it to build something intricate that addresses the big picture. Then, using that information to figure out your next steps before you’ve even turned over the rest of the pieces.

This is predictive analytics and America’s seaports are dipping their toes into the waters, figuring out how to deploy analytics to predict and prepare for infrastructure builds, gate and terminal operations, maintenance and logistical challenges.

Data All Around

Port data can come from many sources. Vessel manifest records are indispensable to seaport activities, including landside cargo operations and throughput billing. But vessel operators also collect data related to their onboard systems, bunkering, provisioning, etc. Port service providers can leverage this additional data into better decision making related to, for example, bunkering.

Information that can ultimately help optimize port operations also comes from cargo management systems such as terminal operating systems, gate logs, crane computers and other sources, or from cargo processing activities, like forwarding, insuring, clearing, inspecting, moving, storing, transferring and billing.

And, data can come from spatial and technical applications. Ports and their partners may collect geological data to be used in construction programs, spatial imagery from position tracking systems, data from radar and optical sensors and security/safety technology, and other technical data. Seaports can use this data to analyze specific assets or patterns of traffic, detect anomalies, and optimize construction or safety and security, among other goals.

Converting Data to Information

Collecting the data might just be the easy part of data analytics. “In order to bring real value to the port’s stakeholders, you must combine deep domain expertise together with superior technology… When these pre-conditions are met – there are limitless opportunities to create value,” said Uri Yoselevich, CEO of DockTech.

Dr. ManWo Ng, consultant and maritime analytics professor at Old Dominion University, said converting data into useful information that is predictive in nature can be tricky, and is dependent on many variables such as the type and format of data. He called predictive analytics an art and a science: “You can give the same data to two of your best analysts and ask them to predict the average truck turn time next week. Their predictions will most likely be different. … Generally, it takes mastery of the technical fundamentals, a healthy dose of creativity and a solid understanding of the dynamics of the maritime industry to come up with useful predictions.”

Analytics at Work

Smart ports are already communicating with partners such as vessel operators (to predict truck congestion), and truck and rail operators (to take steps to reduce landside congestion). Predictive analytics can also optimize labor planning, chassis availability and preventative maintenance. “With accurate predictions, the possibilities are endless,” said Ng.

Of particular interest to seaports is the fact that analytics can cut capital and operating costs.

Yoselevich said, “Predictions exist to empower stakeholders and policymakers in decision-making processes. With the right predictions, a port can optimize its resource allocation – reducing costs, improving handling times and reducing downtimes significantly.

Ashebir Jacob is vice president with global infrastructure advisory firm Moffatt & Nichol. He said that, while no two ports have the same challenges or the same abilities to use predictive analytics for good, every port has opportunities: “Ports have to take into consideration their size of operation, existing data-sharing platforms, union agreements and other considerations. But all container ports, even the smaller ones, can use predictive analytics to optimize their systems.”

The key is to ask the right questions: How can I do something more efficiently? Increase revenue? Improve air emissions? Reduce equipment breakdowns? Enhance safety? Regardless of the goal, “We can use AI to look at data and search for patterns and improvements,” said Jacob. Ports no longer have to wait until, for example, a piece of equipment breaks. “Now we can fix an issue before it becomes an issue,” he said.

In that effort, Jacob said Moffatt & Nichol’s technology division has been perfecting a smart port system that embraces digital twin simulation and emulation. Jacob said that it is important to understand not only port authority dynamics but also those of the greater port community and full supply chain. “Modeling this from beginning to end will help eliminate the real bottlenecks and solve the real problems. We have looked at what our customers are going to need and have developed a solution involving data analytics that will go to market before the end of the year,” said Jacob.

The Challenges

Despite vast potential benefits of port predictive analytics, stumbling blocks remain.

At the Port of Los Angeles, Assistant Marketing Director Chris Chase said, “The hardest part of the process is being able to get good data into the system. The more good sources, the better the models can work.”

At the Port of Cartagena, Manager Salas Trujillo said that the challenges are many, because as the port moves forward it is discovering new focal points for analysis. Project prioritization becomes necessary.

Ng said, “Predictive analytics can be intimidating, but once we strip away all the buzz, it is really nothing to be afraid of.” For example, he worked with one major port to implement a predictive analytics solution using a simple spreadsheet. “That single spreadsheet was enough to cut one cost category of a business unit by 80%,” he said. It is possible to start small, but, he added, “Analytics can and should play a much larger role in solving the current supply chain crisis. It is the next frontier in moving the maritime supply chain forward.”

Jacob said, “Sometimes ports don’t know the extent of their data, or how to make it usable, or what can be done with it.” He added, “Digital infrastructure is different from deepening a channel or building a terminal.” It will take some time for ports to optimize the use of data, but as the parallel e-commerce industry has shown, those that best manage their data will succeed. Jacob said the cost is nominal, but the return from predictive analytics is substantial.

In the Name of Progress

“While predictive analytics has proven to be of tremendous value, oftentimes it is a steppingstone toward prescriptive analytics. Even if you know the future, because of the large number of actions possible, it can be extremely challenging to optimally capitalize on this foresight. This is where prescriptive analytics comes into the picture,” Ng said. It helps identify the best course of action.