From Sustainment to Continuous Improvement: A 21st Century Approach to Support an Operating Fleet

By Robbin Laird

The US has introduced a new generation of combat aircraft.

These aircraft are different from legacy aircraft notably, in that they are designed, crafted and built differently – they are software upgradeable, data rich military aircraft for 21st century forces.

The challenge is to build out the working relationship with industry and to rework military organizations to leverage the opportunities built into the digital rich aircraft.

Leveraging the data generated by a fleet of commonly configured software upgradeable aircraft is a weapon system, which can provide a significant advantage when dealing with peer competitors.

But it will not happen simply by buying an F-35, or a CH-53K.

Much more is required – nothing less than shaping the focus of maintainers, reshaping how the military manages data, and how it works with industry.

During my visit to the Sikorsky Customer Service Center, I addressed this challenge as follows:

As new digital military aircraft systems come into operations, can the military benefit as has the commercial customer from leveraging data to enhance aircraft performance and availability?

 The digital systems in modern aircraft provide for the possibility of continuous improvement. 

 The challenge for the military is to align their organizations with the potential to tap into the aircraft systems THEY ARE ALREADY BUYING.

One of the key persons who briefed me on the functions of the Center as well as its applicability to military platforms was Ping Liu, Chief Data Scientist and Sr., Manager Advanced Analytics, Sikorsky Aircraft.

Given her role in the Center and the insights, which she contributed during my visit, I wanted to follow up with a separate interview with her.

I did so on February 8, 2018 by phone during my stay in Paris, France prior to going to Helsinki.

We started by discussing the role of her team in the Customer Support Center.

“The Fleet Management Room is where my team develops & deploys predictive analytics for the aircrafts in the commercial fleet.

“We want to predict which aircraft is going to be down, and what component we need to stock up to ensure we have the right parts at the right place and right time.

“It is clearly a spiral learning experience.

“When we stood up the Center in 2015, my team was able to apply the latest technologies and capabilities, stand up the fleet monitoring and program performance dashboards so that we can increase the situational awareness across the enterprise.

“Once people understand what’s happening in the fleet, we then identify and integrate some of the best ideas across enterprise, to increase the ability to leverage action-driven information.”

Question: Gaining a fleet knowledge is crucial to the entire effort to understand how to get enhanced aircraft availability.  How do you do this?

Ping Liu: We are shaping predicative understanding of both the fleet and the individual aircraft.

“The fleet is composed of the individual aircraft with individual components.

“But being able to drill-down to the required level and to shape operational insights and convert those insights into actionable decisions is what we are focused upon.”

Question: What about the impact of differently configured aircraft on your ability to understand fleet behavior?

Ping Liu: The S-92 green aircraft has to be configured very differently for different missions. Configuration management is key.

“While we are focused on the fleet performance, we’re also very much focused on understanding regional operator mission differences, so we can actually drive aircraft-specific actions.

“We call it “prescriptive analytics.”

“We need to predict what the fleet is going to do from cost and availability perspective, but also to be prescriptive with regard to each individual aircraft in terms of what they have to do in order to increase performance for that particular aircraft.”

Question: You have regional support centers, and presumably understand the variations in regional operations are a key part of predictive analytics to manage the fleet regionally?

Ping Liu: It is.

“The helicopters flying the North Sea demonstrate different requirements than those flying in the Gulf of Mexico.

“When you are looking at asset deployment especially materiel stocking strategy you have to have a regional focus to be fully effective.

“For example, if  I don’t recognize individual configuration differences, I might be overstocking air conditioning components at North Sea where it’s mostly needed in regions with warmer climate, or overstocking search & rescue hoists that are only applicable to a portion of the fleet in a particular region.

“Understanding regional, operator, A/C level differences are key to manage cost & availability effectively.”

Question: By developing the models, you are obviously opening the door to culling the latest innovations in artificial intelligence and other analytical technologies as well, I would assume?

Ping Liu: That is correct.

“We are using a number of very sophisticated machine learning capabilities to do our job.

“We are consistently assessing the latest AI and machine learning capabilities to increase the development speed and accuracy of our predictive models, so that we understand what the data can tell us.”

Question: When Sikorsky first offered the S-92, it was a platform.  Now you are providing a capability, driven by data empowerment.  Is that a fair characterization?

Ping Liu: It is.

“As more and more data are generated from commercial platforms, we recognize that data needs to be looked at really as a product feature.

“How do we extract value out of this product feature and actually help our customers to improve readiness and cost?

“I think the difference between commercial and military, is on commercial platforms we have more or less easier access to data, invariably, compared to military applications.

“I think commercial operators are incentivized and motivated as they are profit driven, more so than military operations.

“Because of that, anything that can help them to reduce their costs, increase uptime, and they’re usually supportive. And we do a better job therefore delivering value to commercial operators.”

Sikorsky Customer Care Center, Stamford, Connecticut. Photo Credited to Sikorsky.

Question: Let us turn now to the military and how they might reap the advantages of working big data to get enhanced value from their new data rich aircraft.  How best to do this in your view?

Ping Liu: With the K we have an opportunity working with the US Navy and Marine Corps to shape a new way ahead.

“To achieve that we clearly need to have a collaborative data strategy to achieve a breakthrough that the services clearly want and deserve.

“However, this cannot be an after-thought or just an after market strategy.

“It has to be built in from the beginning.

“We have 750+ flight hours so far on the 53K aircrafts. How do we use our early experiences to build up K-specific analytics product and services so that the customer has a much better experience comparing to a legacy system?”

Question: Often when I am talking to military users, they highlight the security challenge but I think what is too often overlooked is that on the commercial side, security is a very high priority from the get go. 
How do you see the challenge?

Ping Liu: It does not matter whether it’s commercial or military.

“We always abide by the most stringent security requirements to protect the fidelity and security of our customers’ data.”

Question: What do you see as barriers for the military to be able to use data as a weapon system, in effect?

Ping Liu: It’s a two part answer, one is data collection, and the other is data quality management.

“For example, on the data collection part the workflow of the maintainers is not necessarily enabling them to handle data in a manner that would allow us to have the kind data that would enable analytics to empower the combat force.

“The Marines are really busy. And if the data collection is too cumbersome, they don’t tend to do it or do it well.

“The Navy can invest in technology and tools to make the data collection a little bit easier.

“For example, when Marines remove a gearbox, they don’t necessarily take pictures of the failed component itself today.

“Also, the database logging the malfunctions is a traditional database with many potentially confusing malfunction codes.

“This may create problems to the Marines while they are trying to describe what has failed.

“In many cases, either they pick a code they remembered but not necessarily the right one, or existing malfunction codes do not describe what they are seeing precisely, so they end up putting additional descriptions/failure symptoms in the comments.

“And the Navy today, lacks the tools to process unstructured data effectively and efficiently, whereas Sikorsky has tremendous capabilities developed over the last few years to do so.

“I think expanding the usage of unstructured data, using text and images to improve fidelity of the data, is something that Navy can do and significantly enhance the ability to gather critical data to improve maintenance efficiencies and increase aircraft availability.

“As to data quality, the data quality management is not necessarily a part of the enterprise work flow.

“But if you truly treat data as a product, the need of quality assurance is therefore a no-brainer.

“There are tremendous opportunities existing in data collection and data quality control which can be shaped to provide for data management which can be a force enabler and multiplier.”

Question: So it is important to think of effective data management, collection and processing and capability as really a strategic asset?

Ping Liu: And if you do so, then you recognize the effort as a priority to ensure that we have a more effective combat force.

“We need to treat data as a product within the military as we do in the commercial world.

“That mind-shift has yet to happen in the military.

“For them to really capitalize on analytics, they have to start to treat data collection, data quality, data management, as a strategic enabling initiative. You need to think about this almost on the same level as lasers, as rockets

“Data needs to be treated as a strategic asset in which downstream analytics are targeted, prescriptive and action-driven to enable enhance force effectiveness.”

In other words, what I would conclude after visiting Sikorsky in both West Palm and Connecticut and watching the standup of the Osprey from the beginning that the K provides an opportunity to not go down the Osprey path of multiple configurations and stove piped data management.

The Marines revolutionized combat by introducing the Osprey, but the configuration management and maintainability side of the equation were not its strong points.

With the coming of the CH-53K there is a strategic opportunity for the Navy-Marine Corps to reshape how they handle data and make it and use it just like any other enabling weapon system.

But again, this will not happen by simply by buying a data rich software upgradeable aircraft.

It requires organizational change and new skill sets built into the force, as well as effective working relationships with industry as well.

This article was first published on March 20, 2018.