This is a guest blogpost by Ali I. Riaz, CEO of OrbitMI
Maritime is one of the most important parts of our global supply chain. Ship owners, ship operators and the entire maritime ecosystem must know where their vessels are at any time, be able to plan routes and see conditions ahead. Maritime transport routing expert Ali Riaz explains how graphs can help.
Planning a route from Port A to Port B sounds simple. But, there are a lot of different components that factor into it, as there’s so much complexity around which ports are sanctioned, the weather situation, avoidance of high-risk areas, having the right insurance policy versus route capability.
From a business requirement perspective, maritime routing is complex to model and optimise. That’s true in the best of all worlds, but it’s especially challenging today.
Maritime operators need to avoid high-risk bottlenecks to minimise unexpected expenses. They need the latest intelligence on any potential port restrictions, and how to route round any such obstacles. More than ever, they need to reach the right port to avoid time spent idling.
They also need to help the drive to Green, so maritime operators want to use routes informed by guidance on ways to limit the danger of oil pollution and associated insurance needs. Ship operators also need to accurately predict the time of transit and consumption through the Emission Control Areas (ECAs), which are new ways in maritime to curb greenhouse gas output.
Only seeing the totality can give you the whole picture
Historically, the sector has relied on older, relational technology to address these challenges. But, this is a complex problem that needs a very reactive, data-driven, real-time solution.
There are multiple, interconnected layers to the datasets for maritime in general —information on cargo, vessels, ports of call, regulations, weather, and location. Every day, the supply chain gets more complex, as more capacity gets added or lost and new traffic patterns emerge.
And that’s not a simple addition as each change ripples out, affecting other data. It’s never a one plus one problem—it’s a ‘plus one plus all the other layers’ situation. It’s only when you look at the totality of the layers that you can see the whole complex picture, and move the needle in this core business process.
Even though 90% of the world’s good are moved via the ocean, I hadn’t seen any other technology in the sector that could cope with as complex a use case as today’s global maritime routing. A custom routing solution was a strategic imperative for us. We knew we had to create a solution that would have to be able to:
- process large volumes of data
- offer reliable storage to house that data and serve it up in real time, and at scale
- support not just linear and tabular but also spatial data sets
- and would come complete with a library of path-finding algorithms.
That meant bringing together the best of artificial intelligence to integrate current and historical Automatic Identification Systems (AIS) data, as well as multiple data feeds via APIs.
We identified a graph database, Neo4j, as the best data building block for the system. That’s because of graph technology’s innate ability to work with complex data structures in a naturalistic way, especially in the way it captures relationships between layers. That gives it clear advantages for this kind of use case over other data approaches.
Our experience shows that graph technology was the only way to create intelligence-based maritime routing capability. Those routing insights mean the difference between success and failure for one of the most critical parts of business itself, the supply chain.