Machine learning transforms predictive maintenance for complex equipment.
of downtime saved in three months
Lockheed Martin revolutionizes aircraft maintenance with the SAS® Platform
The C-130 Hercules is the most versatile aircraft in aviation history. From landing at the world’s highest airstrip in the Himalayas to taking off and landing on an aircraft carrier in the middle of the Atlantic Ocean, the aircraft is celebrated for its unsurpassed versatility, performance and mission effectiveness.
Today, 70 countries rely on the C-130 for search and rescue, peacekeeping, medical evacuations, scientific research, military operations, aerial refueling and humanitarian relief. More than 2,500 C-130s have been produced to date. The worldwide operational fleet includes legacy C-130 models as well as the current production variant – the C-130J Super Hercules.
Because the C-130 is such a vital component of so many global fleets, downtime for any C-130 impacts a customer’s mission readiness.
“The relief efforts are what I’m most proud of with the C-130,” says Mike Isbill, a Lockheed Martin Fellow who specializes in data analytics. “When there was a horrible hurricane in the Philippines, they flew in dozens of C-130s within a day to bring supplies and contribute to relief efforts. You’ll see the same thing around the world with C-130s addressing famine by bringing in food and evacuating people after disasters.”
Lockheed Martin, the C-130’s original equipment manufacturer (OEM), is maximizing uptime with SAS. The company uses artificial intelligence (AI), IoT and advanced analytics to predict when parts will fail, keeping more aircraft airborne for vital missions worldwide.
Our work with SAS has taken us from being a reactive sustainment organization to a proactive organization. Mike Isbill Lockheed Martin Fellow of Data Analytics
Collecting sensor data leads to analytic breakthrough
The journey to insight for complex machinery always begins with data. Thanks to 600 sensors located throughout the aircraft, the C-130J produces 72,000 rows of data per flight hour. Critically, this data includes fault codes on failing parts.
Previously, this IoT data, which streams from the aircraft to the facility that houses it, was stored in different systems. No central repository existed for the more than 400 C-130J aircraft operating in different countries the world. This effectively left engineers at customer sites on their own to maintain the aircraft based on their interpretation of the data, often leading to unnecessary maintenance and multiple days of downtime while parts were shipped from faraway places.
As a child, Isbill loved solving problems. What began with coding video games evolved into a career in analytics. Now, at Lockheed Martin, Isbill set out to offer customers a solution to the problem of excessive downtime.
His first breakthrough came when all C-130J Super Hercules operators agreed to start sharing their flight data. Deluged with information, the challenge then became how to clean and store the data, which came in different formats and initially required three employees to clean each month.
Thus began Isbill’s quest for a robust analytics solution. After a successful proof of concept, Lockheed Martin invested in the SAS Platform and now relies on its high-powered AI, business intelligence and data management capabilities to wrangle the data and build predictive maintenance models.
“The partnership has been amazing,” Isbill says. “With SAS, we’ve reduced data cleanup times by 95%, which has allowed us to spend more time on the actual analytics. With this solution suite and a very small team of data scientists, we have been able to quickly produce results that we believe will reduce downtime and cost for our customers.”
Lockheed Martin – Facts & Figures
on each C-130J
of data per flight hour
in data cleanup time
Intelligent diagnostics via machine learning
With customer data streaming in, Isbill and his team began working on how to use C-130J flight data to profile individual parts and predict when they might break – insight that could help customers proactively stock the right parts and keep aircraft operational for more life-saving missions.
The system uses data from C-130 customers, Lockheed Martin engineers and part vendors to form a central repository on more than 300 aircraft parts. Using machine learning and IoT analytics, the system learns from the collective maintenance history to form a real-time best practice for aircraft maintenance.
For example, if a fault code triggers the replacement of a certain part, which is later found to be in good working condition 80 percent of the time, the system learns from this mistake and next time will recommend more robust troubleshooting before replacement. If the customer rejects this suggestion and takes a different course of action, the system learns from that as well.
Lockheed Martin calls this service “intelligent diagnostics” and believes it has the potential to transform how customers understand and maintain their aircraft.
“We’re seeing a great interest from customers because it reduces their time in getting things done,” Isbill says. “They know what to work on and when. They’re not wasting time trying to figure all that out because we can tell them straight up.”
Monetizing intelligent diagnostics
A future growth area that is of particular importance to Lockheed Martin is offering intelligent diagnostics as an add-on service for customers.
“We’re seeing great results with SAS,” Isbill says. “We’re now starting to work with our supply team to improve not only spares for existing customers but how we sell spares to new customers, and position parts throughout the world for faster distribution.”
One recent project showed promising insights pertaining to the application of intelligent diagnostics related to spare parts optimization. Lockheed Martin worked with one of its largest C-130J operators to track 20 aircraft and 50 parts over three months. Using its predictive maintenance models, Lockheed Martin saved the customer 1,400 hours of downtime – 1,400 hours that could now be used as flying time in support of mission requirements.
Optimizing aircraft maintenance and production
Looking forward, Lockheed Martin wants to expand predictive analytics into other business lines. Of particular interest will be optimizing scheduled maintenance – which is the top driver of aircraft downtime – and reducing production costs.
“Our work with SAS has taken us from being a reactive sustainment organization to a proactive organization,” Isbill says. “We now have insight into the issues that are on the horizon for both us and our customers. By moving analytics into the maintainers’ hands, we are improving the availability for all our fleets, while continually improving our models using the wealth of data we now receive.”
C-130 inspires a career
Mike Isbill said that one of his favorite C-130 stories involves the Vietnam War. In 1975, at the close of the war, 24-year-old Tim Nguyen was on one of the last rescue flights out of Saigon on a C-130 aircraft. In a single flight, that aircraft carried 452 passengers to safety in Thailand. After the escape, Nguyen was so thankful for his safety and so enamored with the rescue aircraft that he committed himself to getting an engineering degree and someday working on C-130s himself. In 1983, he reached his goal when he was hired to work at Lockheed Martin – where he worked until he retired in 2016.
Результаты, описанные в этой истории, относятся к конкретной ситуации заказчика, его бизнес-моделям, исходным данным и вычислительным средам. Опыт каждого клиента SAS уникален и отличается техническими параметрами создаваемой системы, поэтому все заявления носят ситуативный, а не общий характер. Фактические результаты, экономия, производительность и изменения в ключевых показателях эффективности могут варьироваться в зависимости от конфигурации решения и бизнес-условий каждого заказчика. SAS не гарантирует и не утверждает, что каждый заказчик получит такие же результаты, как описаны здесь. Единственными гарантиями для продуктов и услуг SAS являются те, что заявлены в письменном соглашении по соответствующим продуктам и услугам. Ничто из описанного в данном материале не может расцениваться как дополнительные гарантии. Заказчики поделились с SAS своими достижениями и результатами в соответствии с условиями договора или после подведения итогов успешного внедрения программного обеспечения SAS. Наименования продуктов являются торговыми марками соответствующих компаний.