The US NGAD and Digital Engineering: Lessons for Pakistan

Author Profile: Syed Aseem Ul Islam is PhD candidate at the University of Michigan, Ann Arbor, USA, specializing in adaptive and model-predictive flight control systems. He received his bachelor’s degree in aerospace engineering from the Institute of Space Technology, Islamabad, and his master’s degree in flight dynamics and control from the University of Michigan.

As combat aircraft have gotten complex (especially with the progression to 4+ and 5-generation designs), their lifecycles have become much longer. The wearily long development cycle of the F-35 is well known to many. The development of that fighter began in 1995, yet its first pilots began training on the program 17 years later in 2012. Likewise, the F-22 – which began as the Advanced Tactical Fighter (ATF) in 1981 – took 22 years to its first production delivery.

Overall, since a country can only fund a finite number of high-cost, high-complexity programs at the same time, the current crop of new aircraft must fulfill multiple roles. Thus, they are equipped with the most cutting-edge systems, which drives up complexity and cost.

Furthermore, because of long development cycles, the eventual end-user requests new systems on top of the initial set of specifications, which leads to delays. In fact, this has gotten more common due to the increased rate of technological advances in recent years.

The long development cycles, coupled with delays, inevitably result in cost overruns. Despite that reality, because these projects are high priority – and with large amounts of money spent already – they become “too big to fail,” and thus, compound the cost overrun issue.

It is a classic case of a snake biting its own tail; these overruns raise the stakes of the project and, in turn, lead to even more exotic specifications to justify the cost and delays.

Finally, since the development costs inevitably balloon, one must distribute the costs over a larger number of aircraft. In turn, the end-user would have to keep these aircraft in service for a longer period to keep the life cycle costs at a justifiable level.

However, the induction of many aircraft is not always possible. Moreover, keeping these aircraft in service for an extended time-period necessitates an expensive service-life extension program (SLEP).

With these realities in mind, the Assistant Secretary of the United States Air Force (USAF), Dr. Will Roper, proposed and implemented a radical paradigm shift in aircraft life cycles. Roper proposed the “Digital Century Series” model. The name hearkens back to the group of six F-1XX fighters and fighter-bombers produced by multiple aircraft manufacturers for the purpose of meeting a variety of distinct roles in the USAF in the 1960s. Developed quickly, these were relatively simple aircraft – and they did not serve for decades. Digital Century Series is a modern revisit of this concept.

Digital Century Series Acquisition Model

Developing several aircraft over short timeframes is the main outcome of this acquisition model. With the use of digital methods, the hope is to develop relatively low-cost aircraft quickly and, in turn, retire them sooner (than their 4^th and 5-generation predecessors), potentially at 3,500 hours. Basically, the end-user would replace its older aircraft with a new platform every eight years. Clearly, this is a radical departure from the current norm of operating fighters, which is to use them for decades at a time.

These are key components of the Digital Century Series development and acquisition doctrine:

Planned Obsolescence:

The U.S. will only design aircraft with a lifespan of 3,500 flying hours, and it will not initiate SLEP programs to extend their service lives. This could reduce life-cycle costs by avoiding costly SLEPs and, possibly, help increase aircraft operational availability. Basically, the U.S. could procure batches of 50-80 aircraft of each new type every eight or so years. In addition, the model will ensure that the aircraft in service boast only the most cutting-edge systems available and, in turn, are not limited by legacy technology roots.

Multiple and Simpler Types:

The U.S. will generally design each aircraft for specialized roles. Basically, it will not beset one or two types with air-superiority, interception, suppression of enemy air defence (SEAD), strike, and electronic warfare (EW). Rather, it will tailor specific platforms for each of those roles. This will allow the U.S. to design each aircraft on a simpler basis and, as a result, drive development and operational costs downward. This would also allow multiple competitors to compete and drive up design quality. The hope is that the high-volume of design work will lead to more innovation, as opposed to volume in production.

Commonalities:

Each aircraft type will also draw on a common set of subsystems, such as ground-service equipment (GSE), software interfaces, human-machine interfaces (HMI), and even some structural components, such as the landing gears and engines/powerplants. This will reduce development and operating costs as the industry and service can spread core input costs over the long-term and over large numbers of applications.

Digital Engineering and Digital Twins:

Under Digital Century Series, the U.S. will design and develop aircraft using digital engineering concepts that should greatly shorten development timelines. Moreover, the end-user will maintain digital twins of aircraft that will cut maintenance costs and boost aircraft availability through predictive maintenance.

Next Generation Air Dominance (NGAD)

On 14 September, 2020, Will Roper announced that the industry designed, built, and flown a full-scale flight demonstrator of the ‘Next Generation Air Dominance’ (NGAD) all within a little over a year.

Roper did not disclose any specific details about the NGAD, but the announcement surprised observers as 5^th-generation fighters were only entering widespread service. The flight of a 6^th-generation demonstrator is certainly impressive. As emphasized by Dr. Roper, the details of the aircraft are not important; rather, one must focus on the fact that the digital engineering model evidently works in the real world, i.e., it can be a means to shorten and cut the cost of the development cycle.

Digital Engineering and Digital Twins

The ability to rapidly develop and fly the NGAD demonstrator boils down to the use of digital engineering. To design, develop, build, and fly an aircraft in such a short time is unfathomable with conventional design methods. This begs the question: what is digital engineering?

The key driving digital engineering is the idea that computational power has grown to a point where it is feasible to supplant time and resource-intensive legacy physical tasks in aircraft development with high-fidelity simulations that can yield results in a fraction of the time. In fact, the industry is already using new computer-based tools to develop contemporary aircraft, but NGAD had shifted the full development cycle to the digital domain.

Central to the process was the creation and use of a digital twin. In general, the development of a digital twin encompasses the following three main stages:

1. Digital Twin Prototype (DTP)

Using knowledge of physics and historical data, the developer designs and models each component of an aircraft in a computer. In fact, they would model components down to the individual cables, the switches, the actuators, and the structural members.

These computer models allow the producer to test the aircraft in numerous ways without manufacturing anything physically. They can even model processes – such as casting and machining – computationally. Furthermore, they can simulate life-cycle phenomenon such as fatigue over thousands of cycles on a computer in short timeframes, all without the need to manufacture a single input.

This process enables the producer to fully refine component designs before they push those parts to manufacturing for the first time. The increase in designing and prototyping speed this approach can provide is nothing short of revolutionary as it significantly reduces the number of physical prototypes one would need for development and testing.

2. Digital Twin Instance (DTI)

The DTI is the validated version of the DTP model. One validates the DTP by physically manufacturing the component for the first time, verifying that the model is correct, and refining the model using data from the physical component. The producer uses machine learning methods to refine DTP models with physical data to produce DTI models of each component. Thus, DTI models merge knowledge of physicals and real-world data to produce a superior model. Due to their nature, DTI models are extremely accurate and offer significant predictive capabilities.

3. Digital Twin Aggregate (DTA)

This is the final step in the digital twin creation process, and is the name given to the aggregation and interconnection of all the DTI models to produce a single, extremely high-fidelity model of a complex system, such as an aircraft, called the “Digital Twin.”

In the context of aircraft, design, manufacture, and maintenance, digital twins play a crucial role. As shown with the NGAD test flight, one can drastically speed-up the design process. This rapid development cycle is impossible with traditional design methods, which rely on several prototypes of each component in the system. In manufacturing, digital twins can help streamline manufacturing processes before the team can make expensive mistakes. Moreover, one can specify machining and casting processes, design the jigs and optimize production lines to increase speed, cut costs, and raise quality.

In terms of maintenance, the biggest advantage of maintaining digital twins of aircraft lies in the realm of predictive maintenance. As each aircraft rolls off the production line, the system creates that unit’s digital twin. So, as the aircraft completes sorties, the data from those flights feeds into the digital twin, which in turn uses the data to accurately age the digital twin to reflect the current state of its real-world mirror.

The digital twin can enable the maintenance crew to identify potential issues before they arise. The user can use individual parts well beyond the stated/nominal design life if those parts have more life in them. This can help reduce costs, increase the efficiency of maintenance, and push high aircraft availability rates by minimizing downtime.

Lessons for Project Azm?

From the onset, Pakistan Aeronautical Complex (PAC) cannot – and should not – try mimicking the USAF’s approach. However, it can learn important lessons from the U.S.’ use of digital engineering, and the Digital Century Series fighter designs, but work within Pakistan’s constraints and realities.

Unlike USAF, the Pakistan Air Force (PAF) cannot afford to operate six or more new types of fighters every eight years. However, the PAF already operates three distinct types, i.e., a hi-lo combination of the F-16s and JF-17s, and a limited number of Mirage III/5 for stand-off range weapon (SOW) deployment.

Thus, the PAF can create its own version of the Digital Century Series doctrine in the future by splitting its Air Staff Requirement (ASR) across distinct – and achievable – requirements. It can, for example, build off its current framework, i.e., a medium air superiority fighter (F-16), a lightweight multirole fighter (JF-17), and a deep-strike-optimized attack fighter (Mirage III/5). In turn, the Pakistani industry can continuously replace aircraft in these tiers every 10-12 years. If Pakistan’s aerospace sector progresses sufficiently, one can hope for local designs and inputs to support each ASR by the 2050s or 2060s.

The most important lesson PAC can learn is one of commonality and modularity. For example, PAC could develop critical inputs – e.g., flight management system, the flight control system, HMI, actuators, and sensors – and use them onboard multiple platforms. PAC could even make these inputs available to the private sector and, in turn, empower Pakistani companies to rapidly develop a wide range of solutions for the PAF. The PAF could leverage many solutions, but also boost interoperability and cut operating costs.

By setting up the infrastructure for digital engineering, PAC will also have an easier time managing parallel projects by efficiently leveraging hardware, software, design, and techniques. So, for example, it can run several unmanned aerial vehicle (UAV) projects – e.g., a loyal wingman UAV and reconnaissance UAV – by employing common resources and cut costs through rapid development.

It need not set-up parallel jigs to build, test, and iterate prototypes while developing the designs. Overall, the software, data and models will all be common across multiple platforms, and available for testing and refining digitally. This will result in a dramatic increase in design and development speed.

Yes, PAC is a relatively small and much newer entrant in the aerospace industry, and it lacks capacity and experience. However, it is also free from the massive institutional inertia and deadweight that could stop it from even exploring an equivalent to the Digital Century Series, much less employ it. PAC can move to new engineering and production processes right from the onset.

For example, the establishment of the Centre of Artificial Intelligence and Computing (CENTAIC) – when AI/ML adoption in defence is still in its infancy – is a promising sign. The PAF is evidently interested in the adoption of new technology concepts, so studying and implementing digital engineering is plausible. It is worth noting that PAC is using 3D models on the assembly line of the JF-17 (instead of legacy drawings).

3D models seen on the JF-17 production line in a recent program aired by Hum TV.

The PAF has always been a force that has strived to maintain a technological edge over the IAF, as a numerical edge is out of the question. Often being cash strapped, the PAF will be able to reduce costs, and increase aircraft availability as well. The adoption of digital engineering and Digital Century Series-like acquisition doctrines may lead to a revolutionary change in PAF’s capabilities.