Follow Datanami:
September 25, 2018

Cat Gets Smarter at Sea

(RomanLebedev/Shutterstock)

Caterpillar’s marine business has integrated a suite of machine learning software from OSIsoft that should help its clients save millions in fuel costs while improving the safety and reliability of equipment on the high seas.

OSIsoft announced today that Caterpillar is OEMing some of its technology, including the Connected Services and PI Systems offerings, as part of Cat Asset Intelligence, which is a suite of machine learning-based services that Caterpillar offers to clients in various industries.

The complexity and scale of today’s ships make them good candidates for optimization through big data. According to Matt Miller, OSIsoft’s transportation industry principal, the two most immediate areas for savings are energy consumption and maintenance costs.

“Ships use tremendous amounts of fuel,” Miller tells Datanami. “A fully loaded tanker can burn 250 tons of bunker fuel a day. Over the lifetime of ship, fuel can be 50% of the operating costs. Reducing it by even a few percentage points can mean hundreds of thousands of dollars or more per ship per year.”

OSIsoft and Caterpillar have collaborated with one shipping customer that runs a fleet of RoRo ships, which are a type of large cargo ship. Miller says that CAT Asset Intelligence, combined with data flowing from OSIsoft’s PI System, showed the company how marine life growing on the hull was creating drag and increasing fuel burn.

CAT Asset Intelligence can analyze data from engines made by Caterpillar and other manufacturers

“After revealing the program CAT Asset Intelligence came up with a schedule for optimized cleaning,” which should save the company $450,000 in fuel per vessel per year, he says.

Another marine customer that operates inland tugboats was alerted to problems with its diesel engine from a third party manufacturer. According to OSIsoft, by analyzing a combination of fuel pump supply pressure, engine performance, and other data, the tugboat operator was alerted to a fuel pump problem before it became critical.

“Fixing something before it breaks can reduce the costs easily by 50% and often it’s more,” Miller says. “If a part breaks, and there’s nothing in port, the cost associated with lost time compound the costs you’re already incurring in repairs.”

According to Miller, the CAT Asset Intelligence solution incorporated 40 million hours of equipment analytics experience across more than 80 OEM partners, including manufacturers of other diesel engines. OSIsoft specializes in turning streams of data from a wide variety of sensors into a format that’s usable by humans.

“We developed over 500 connectors that pretty much allow us to get data from any industrial device you can throw at us,” Miller says. “We even see equipment from the 70s, which I’m sure you can imagine presents challenges to data access. Second, I’d say it’s that emphasis on the ‘human factors.’ We structure the data and serve it up so that people can start solving problems.”

Turning sensor data into operational insight is “a far more difficult task” than most people would think, especially for OSIsoft customers that are tracking 20 million or more data streams at once. “You have to serve it up in ways that people can understand and see problems right away,” Miller says. “It won’t do you any good to know an accident is about to happen ten minutes from now if it takes two days to get an answer. We’re almost like an EKG machine for big operations.”

Related Items:

From Big Guns to Big Data: US Navy Looks to Tech for Advantages

How Big Data Can Optimize Global Shipping

 

Datanami