Bell Flight is the company behind some of the most iconic and groundbreaking aircraft of the 20th century. As depicted in the film, “The Right Stuff,” US Air Force test pilot Chuck Yeager was the first to break the sound barrier in the Bell X-1. And Bell Flight also designed and built the world’s first commercial helicopter, the military medivac version of which is seen in the opening sequence of the hit TV show M*A*S*H.
Today, 85 years after its founding, Bell Flight is still positioned on the leading edge of aviation through its development of tiltrotor aircraft. Owing to their versatility—combining the vertical-lift benefits of helicopters and the speed of fixed-wing planes—the US Department of Defense has fielded tiltrotor aircraft and is evaluating further application as the US Army prepares to replace the Black Hawk.
A new dimension of aircraft performance
There’s a common understanding among all defense contractors that the future will bring a new kind of warfare, one that requires a new generation of weapons that are smarter, more mobile and more autonomous. But there’s also another dimension of weapons performance that we see playing a stealthier role on the battlefield: using advanced analytics to improve readiness.
Within the group I manage in Bell Flight’s Future Vertical Lift program, called Digital Systems, we have a team focused on using aircraft sensor data to fundamentally change how the US military keeps its aircraft fleet at maximum readiness. From the perspective that counts the most—that of military leaders sending aircraft with soldiers into forward operating areas—readiness means that when the chips are down, the equipment will perform. Anything else puts the mission and lives at risk.
So where does data come in? Even though a piece of military hardware can appear combat ready, there may be parts or components on the verge of failure, and that means risk hiding in plain sight. The core of our vision is that artificial intelligence (AI) technologies like deep learning can find—in fact predict—those risks before they affect performance and combat readiness.
Predicting problems, proposing solutions in aircraft life cycles
Also critical are the costs of supporting the next generation of combat aircraft, some of which are expected to be in service more than 50 years. We recognized that with operations and support accounting for some 70 percent of a fleet’s total product life cycle costs, the ability to use data to predict problems has the potential to drastically lower those costs.
When we engaged IBM to discuss our options, their team laid out a similar project IBM had done for the US Army for armored fighting vehicles. As part of the project, the Army provided IBM with 15 years of maintenance history on a fleet of 350 armored fighting vehicles, along with roughly 5 billion sensor readings from the same period. Using deep-learning AI algorithms and IOT technology like digital twin, the IBM Equipment Maintenance Assistant solution created predictions of which vehicles were in need of preventative maintenance.
Then, by feeding a wide range of technical manuals into that pool of data, the system was able to tie each failure prediction back to a root cause, down to the part or component level. That made it possible to automatically formulate detailed recommendations for Army aircraft maintainers to follow. One powerful benefit illustrated by the exercise was how predicting the failure of a relatively low-cost part can avoid the need to replace an entire system—such as an engine—saving well over US$12 million.
Faster to the field
The power of this example was one reason we are working with IBM Global Business Services – Asset Optimization Services as an asset performance management provider. Another was the sheer breadth of experience IBM brought to bear in the area of US Army logistics systems, which compliments the value proposition of Bell Flight’s future vertical lift programs.
Furthermore, by integrating predictive insights into our ERP systems, we have the potential to vastly improve the way parts inventories are managed. Take parts defects, for example. Detecting them early is critical to avoid overpredicting the number of available parts, which can easily threaten timely manufacturing and delivery. This integration of the IBM IOT solution with an ERP system is an example where we could detect problems before they can have an impact on our manufacturing runs and delivery dates, thus speeding the availability of our aircraft in the field.
Improved aircraft readiness through analytics
With more transparency into underlying part failure patterns, the Army has the potential to better manage its spare parts inventories, and further translate those insights into improved budgeting and maintenance management practices going forward.
In the same way, we see the solution making it easier and more cost effective for Army maintainers to keep their aircraft in action. In terms of metrics, we believe that the IBM IOT solution has the potential to increase the mean time between failure for US Army aircraft, while reducing the mean time to resolution for problems and promoting longer maintenance free operating periods.
It’s a combination that translates into improved combat readiness and a reduction in mission.