The MarineInsight platform from US-based ioCurrents uses machine learning to monitor on-board equipment and pre-empt otherwise costly failures. Brian Dixon looks at the company and its service for the Fathom World Startup Profile: 2020 series.
There is much to be said for the old adages that ‘a stitch in time saves nine’ and ‘prevention is better than cure’. What’s more, thanks to the MarineInsight analytics platform developed by Seattle startup ioCurrents, maritime industry users are now not only able to put these pearls of wisdom into practice, but also reduce costs while simultaneously improving safety and performance.
“Predictive analytics enable maritime operators to move to a more reliable, efficient and proactive mode of operation,” says CEO and co-founder Cosmo King. “Instead of responding to alarms when they become critical, operators can service equipment in advance, often avoiding unplanned downtime entirely. Using real-time data analysis on board the vessel, operators can find ideal operating modes for fuel efficiency and adjust in real time.”
“They can also see the implications of behaviours like excessive idling on both fuel wastage and on engine performance and maintenance,” he continues, adding that data models “get better over time and provide safer and more economical methods of operation than can be provided by the equipment manufacturers’ generalised specifications”.
To achieve all this, the MarineInsight platform combines two main components: an on-board mini-computer dubbed the DataHub that collects and analyses data locally using edge computing; and a secure cloud platform. “The DataHub gathers data from a variety of on-board equipment, such as engines, generators, Z-drives, tanks, winches, [programmable logic controllers (PLCs)] or alarm panels,” King explains. Meanwhile, the cloud platform “creates models to represent normal operating metrics for that specific equipment and with respect to the greater fleet”.
The platform then analyses these models to detect anomalies in real time, sending out notifications to the user should any such deviations be determined. “Encrypted data is synchronised from the DataHub to the cloud platform through a cellular or satellite connection to provide the same level of visibility for the vessel and shoreside team,” he says. “Reports and summaries of the analysis are available throughout the organisation so that maritime companies can continuously understand and improve the operation of the fleet.”
“ioCurrents is the only independent technology company providing machine learning-based maintenance advice for the maritime industry,” King states. “We believe our independence better serves operators and we have a system that adds new applications and grows with them over time. Also, when it comes to analytics, edge analytics is essential. Analysing operational data on board the vessels adds immediate insight and drastically reduces the bandwidth, [resulting in] more speed and more security, which is needed to transmit quick answers to the company. The other alternative systems using traditional [Internet of things (IoT) technology] cannot work on a vessel given the large volume of on-ship data and the limited bandwidth [and] intermittent communications.”
At present, MarineInsight is being used to serve “fleets of vessels across the maritime industry”, ranging from inland river towboats and commercial fishing vessels to bulk cargo ships and tankers. “In fact,” King says, “our early products were developed and tested for some of the harshest environments with commercial pollock trawlers in the Bering Sea off Alaska and Western Towboat, [which] hauls freight offshore between Seattle and Alaska in very challenging conditions.”
“With Western Towboat, we were able to go from multiple unexpected main engine failures a year to none in three years. Today, we are working with some of America’s largest towboat fleets to improve maintenance on each vessel and compare performance of vessels across the fleet,” he continues, adding that ioCurrents has also recently signed a number of new contracts within the US shipping and towing industry. Meanwhile, in terms of its offering, the company, King reveals, currently has “several exciting product releases” lined up for the near future.
Having spent “a lot of time in big data companies in the tech industry, such as Isilon Systems, Microsoft, EMC, VMWare, Hitachi and SpaceCurve”, King and CTO Bhaskar Bhattacharyya co-founded ioCurrents in April 2015 after finding that “existing OEM-developed technology solutions were not sufficient for the maritime industry”. At the same time, they both saw that there was “a huge opportunity to increase safety, reliability and transparency by building a platform for real-time analytics of ships’ operational data”. In order to be successful, though, “they felt that a new business model was needed”, one that “emphasised independent, unbiased analysis as a service”.