15 Apr

AI-enabled sensors 

Shipping companies have shown growing interest in internet of things (IoT) for equipment on their assets to remotely monitor machinery, ship condition and performance. 

This is part of the industry’s digitalisation drive and an aspect of the sector’s sustainability strategy. 

However, there are limitations to the amount of data that can be transferred from ships to shore and on the return route. 

Maritime communications providers have made great gains in recent years to ramp up bandwidth capacity and segregate ship data transmissions from general vessel communications.

 But there remain capacity constraints amid continuously rising bandwidth demand. One solution comes from developing sensors with AI capabilities. 

These will process data at source before transferring useful information to a central collection server on a ship’s bridge or engineroom control centre. 

Lux researcher Cole McCollum highlighted the potential of AI-enabled sensors in a December-published report.

“Recent advancements in machine learning capabilities enable developers and operators to extract more value out of sensors,” Mr McCollum said. 

“This is an opportunity to create new products and improve internal processes by generating deeper insights off existing hardware.” 

Shipping companies could retrofit existing machinery on vessels with AI-enabled sensors to generate data streams for condition and performance monitoring

Data would be processed on the sensor and only relevant information or changes in performance or condition would be sent to the central unit and via satellite. 

Mr McCollum urged companies to “consider existing sensor deployments and examine how more value can be wrung out of the data.”

Start-up companies such as Algorithmica are developing solutions that can generate more insightful data from lower-cost sensors. There are other applications in maritime sectors for AI-enabled sensors. 

Mr McCollum suggested “AI-enabled sensors are powerful when coupled with automation, as in 3D printing and robotics.”