Ships are the beating heart of commerce, responsible for moving 90% of global trade, according to the International Maritime Organisation. They are on the move constantly and as the world becomes more and more connected, our reliance on them in our daily lives only grows further.

The maritime industry might be one of the world’s oldest and most enduring, but it’s not always the most technologically advanced, relying on extensive paper logs while other industries raced to digitise. 

More recently, satellite internet innovations have enabled fleets beam reams of data to databases, but the actual use of that data has been minimal.

That’s a missed opportunity.

Inefficient manual processes and a lack of predictive maintenance can lead to avoidable machinery breakdowns, subpar fuel consumption and elevated exposure to operational risks. All of this makes maritime an industry ripe for innovation and transformation by technologies like Artificial Intelligence.

“The components on vessels, from engines to generators, could have hundreds or thousands of sensors, some pooling data hundreds of thousands times a second, creating gigabytes and gigabytes of data every day. But historically this data has been woefully underutilised. It runs around the local network of the boat then disappears into the ether,” says Christian Harrison, one of the founding engineers at Ceto, a tech startup aiming to reshape the future of maritime, which just announced a $4.8 seed round.

“There are two key problems that have existed. Firstly, a lack of adoption. 

Historically people haven't realised how valuable that data can be and there are far too many data points, deeply hidden patterns and layers within the data sets for a human brain to get their head around,” he tells The Stack.

He adds: “If you have a vessel in the middle of the Atlantic Ocean producing gigabytes of data, previously there was no way to get that data off the boat, especially live. By the time it’s ashore, you've got reams of it and people don’t know what to do with it. Previously we lived in an informational drought – now we live in an informational flood but no one can swim.” 

Riding the waves

Ceto was founded by Tony Hildrew, who saw these issues first-hand in the engine rooms of ships and then as a fleet manager ashore. It captures and analyses high-frequency data to prevent machinery breakdowns on vessels. It also provides the first connected marine insurance policy. 

Named after the Greek goddess of the sea, Ceto uses AI and machine learning models to detect patterns, similarities and anomalies in the data collected from sensors on ships. The ultimate aim is to preempt failures, streamline operations and manage risk more proactively. 

See also: "Everyone was laughing: Now they take us more seriously..." Europe’s biggest retailer turns cloud provider

“We’re trying to bring maritime into the modern data age, de-risking, decarbonising and creating efficiency gains across the sector,” says Harrison. 

“Our job is to teach them how to swim with data. We are taking the data and unlocking value, be that behavioural stuff with telematics, hitting emissions targets, predictive maintenance or for insurance purposes.

“From a predictive maintenance perspective, our Mechanical Performance Index (MPI) is like a credit score but for assessing the risk profile of a vessel. Off the back of our MPI score, our decision systems alert the end-user that there might be some things that requires attention.” 

Abandon ship

Developing the technical backbone of Ceto’s platform was a journey. Initially, Ceto relied on InfluxDB to manage the data streaming from sensors on ships. 

While this was fine in the early days, when a single IoT device was plugged into one component on a small vessel, Ceto quickly ran into scalability and reliability issues. The InfluxDB system was overwhelmed with both the volume and velocity of data, with fleets generating hundreds of megabytes of raw data per vessel per day. Inefficiencies and downtime resulting from the system’s struggles to cope negatively impacted Ceto’s business.

“We saw what the industry was doing and we knew we had to move quickly to stay ahead,” says Harrison. “We needed a one-stop shop. We needed to be able to have all of our time series data, all of our user data, all of our vessel data, all of our models, everything in one place, queryable in the same query language. No one has to go elsewhere for their needs.” 

One stop shop

To realise its ambitious vision for the maritime industry, Ceto eventually landed on MongoDB Atlas as its managed database solution, specifically MongoDB Time Series Collections. This was a defining moment, Harrison tells The Stack, as it rapidly accelerated Ceto’s capabilities.

The benefits have been plentiful, not least dramatically reducing data storage needs from 300MB to 3MB per vessel per day due to MongoDB’s superior data compression capabilities. Greater predictive maintenance capabilities have reduced downtime and extended machinery lifespan, saving over $20,000 per vessel annually, and data analysis improvements have increased customer satisfaction and operational reliability, enabling Ceto to enhance its SLAs.

Today, Ceto is managing data for some of the biggest shipping companies in the world, with approximately 35 vessels currently in its pipeline.

 The IoT device is now a much more mature technology and there are also satellite devices around each vessel feeding into a central hub. 

“We're data hungry and MongoDB Atlas has just been our perfect one-stop shop for everything we need data-wise,” Kennedy adds. “One of my first jobs when I joined the company 18 months ago was migrating from InfluxDB to MongoDB, which was a pain but once done an absolute pleasure. 

“Working with MongoDB is a no-brainer. No hosting requirements, no scalability issues. As a small team that for the most part only uses MongoDB, anyone can jump into anyone else's work or their departments and understand the flow. It's the same query language. It's the same process.

“We're a small startup company that’s growing quickly. We don't know where we'll be in 12 months or even six months, and one of the things we love about MongoDB is that so much of the hard work is already done for us. We do some stuff, but occasionally MongoDB just flicks a switch behind the scenes and all of a sudden we have gained a 100% improvement.”

Delivered in partnership with MongoDB

The link has been copied!