Lockheed Martin, in partnership with SAS Industries, is leveraging machine learning and artificial intelligence (AI) to revolutionize aircraft maintenance and performance to ensure the C-130J Super Hercules is ready for what’s next.

A dedicated team developed a tool suite called HercFusion, which uses data from almost 3 million C-130J flight hours to predict when parts will need to be replaced. HercFusion keeps fleets airborne in support of missions around the world.

What it Means

HercFusion provides C-130J operators with machine-learning results that inform maintainers in how to better maintain the aircraft, resulting in:

  • Improved availability of the aircraft
  • Increased mission capability of the aircraft
  • Reduced cost over time

“HercFusion allows the maintenance ops team and the flight ops team to look at the health of an aircraft, down to the part level, and determine the best aircraft to deploy,” said Mike Isbill, Lockheed Martin Technical Fellow who specializes in Digital Sustainment Analytics. “That way, users can schedule when they do their maintenance while they have all the parts and support equipment in place to do that [maintenance].”

HercFusion brings 5th generation sustainment to the user, allowing operators to shift from unscheduled to scheduled maintenance.

Data Matters

Each C-130J Super Hercules has 600 sensors that generate 3GB of data per flight hour.

HercFusion analyzes all of that data and uses algorithms to predict when a part needs to be replaced so that maintenance can preposition spares to keep aircraft up and running.

These insights allow customers to better plan for deployments.

“It lets operators know what they need to take in their pack-up kits, because they know the health of that aircraft when they get ready to deploy,” Isbill said.

By the Numbers

This predictive maintenance model shows a 3% increase in mission capability rate.

  • “That may seem like a small number but it actually can represent having a completely extra aircraft in your fleet,” Isbill said. “It’s huge cost savings, it’s an aircraft they didn’t have to buy, it’s parts they don’t have to buy. We’re getting more up-time for the customer, at lower cost to them and a safer aircraft for the crew.”

One HercFusion operator reported about a 15% reduction in fuel usage.

  • “We’re able to reduce some of their maintenance time, and they’ve actually seen about a 15% reduction in fuel usage, so a cost savings, saving to the environment, and the goal is to continue to improve that mission capable rate,” Isbill said. “The less time the aircraft is down having to do maintenance — especially if it’s troubleshooting you really don’t need to do because our AI can tell you don’t need to do it — is a huge benefit.”

What’s Next

In an ever-evolving battlespace, our team continues to leverage AI technology to help our customers complete missions with enhanced speed, accuracy and safety. Now, the team is working to compact the machine learning and the AI tool to actually go on the aircraft.

  • When aircraft fly away, the operators will take that tool with them. Our team will then feed the data back to the base using 5g.mil, where the maintainers on base will be able to review it in almost real-time.
  • “You’re not going to have bases in permanent places; you’re going to have bases that have to move. You have to be able to know where the parts need to be, when [the parts] need to be there, and then get [the parts] on and out as quickly as possible,” Isbill said.

By adjusting the data and operating environments, these algorithms can work for any aircraft and the team aims to apply this machine learning and AI to other platforms and products.

  • “Every day we move forward, we create new algorithms, we create improvements to the algorithms we have,” Isbill said. “We get closer and closer to getting that downtime to where it’s just removing the parts you need to remove, put the new one on and go.”

Release Lockheed Martin
Photo Rob Vogelaar

Bir yanıt yazın

E-posta adresiniz yayınlanmayacak. Gerekli alanlar * ile işaretlenmişlerdir