九州体育

Smarter skies: How AI and open architectures are changing military avionics

Story

May 07, 2025

Fixed-wing aircraft avionics systems, once the home of gauges and dials, are now resembling cockpits imagined by science fiction authors. They are not just adopting digital technology but embracing artificial intelligence (AI) capabilities and super-performing signal processors.

Multicore processors now crunch data many times faster while consuming a fraction of the power. AI is moving from concept to cockpit, helping manage cognitive workload for pilots facing complex battlespaces. Additionally, open architecture approaches are breaking down proprietary barriers throughout the avionics industry, enabling new levels of integration and upgrade speed.

Major questions remain, however: How can industry build systems that balance automation with human oversight? What’s the best way to protect increasingly connected aircraft from cyber threats? As avionics become more sophisticated, how can designers ensure pilots aren’t overwhelmed by the very systems meant to help them? These are the questions driving the development of military fixed-wing avionics today.

AI and computing advances

The biggest changes industry sources have noticed in the past year in this area are faster computing, AI integration, and smarter systems that enhance combat effectiveness while maintaining strict size, weight, and power (SWaP) requirements.

Collins Aerospace (Charlotte, North Carolina) engineers are are developing “capabilities in sensing and situational awareness, resilient navigation, and pilot-vehicle interfaces that benefit from these advances in open system architectures and advanced computing,” says a company spokesperson.

AI adoption is driving many developmental efforts in the area of avionics.

“The next wave of avionics technology includes the integration of AI and machine learning (ML) tools to create new functions in avionics,” says Pratish Shah, U.S. general manager at Aitech (Chatsworth, California). “For example, AI and ML can be used to better process data to deliver information to the crew, improve in-flight decision making for increased safety and performance, or for better support of warfighters. These are capabilities that we can expect to see in production avionics in the near future.”

Implementing AI in avionics systems remains technically challenging. “The process of integrating AI and avionics systems is complex due to the power- and thermal-management requirements for current processors,” Shah explains. “In order to successfully bridge AI and avionics systems, there’s a need for AI coprocessors that are highly SWaP-efficient.”

Fortunately, he notes, “advancements in SWaP-efficient AI co-processors are helping to enable the adoption of AI into avionics computers.”

Dealing with information overload

AI is also seen as a tool for addressing the growing problem of information overload in cockpits.

David Slack, director of engineering at Times Microwave (Wallingford, Connecticut), says he sees pilot sensory overload as a persistent challenge that’s becoming more acute. “Artificial intelligence is being implemented to mitigate pilot workload by aiding in data analysis and decision-making,” he says. “[This] simplifies complex data streams, enabling pilots to make faster and more informed decisions.” (Figure 1.)

[Figure 1 | Times Microwave’s PhaseTrack cable assemblies use a proprietary dielectric to maintain stable phase performance across temperature variations.]

Ryan Walters, general manager of Thales Flight Avionics USA [Arlington, Virginia], says this capability is increasingly important in today’s combat environment.

“As you can imagine, contemporary and future aircrews will probably not have institutional knowledge or combat experience previous aircrews had, so how do we help them manage a very stressful high workload environment?” he asks. “We see AI helping with cognitive workload mitigation, making critical decisions with humans in the loop.” (Figure 2.)

[Figure 2 | The Thales Scorpion head-mounted display uses augmented reality and hybrid inertial-optical tracking to improve targeting and situational awareness across fixed- and rotary-wing platforms.]

These AI systems aren’t just handling single tasks but are being developed to manage complex scenarios with multiple elements.

“AI can come into the picture and really manage the amount of stuff that’s happening at once,” Walters adds.

Aitech’s Shah points out that AI-enhanced systems could eventually lead to significant automation of flight functions. “With AI and ML tools, semi-autonomous, and eventually fully autonomous, decision-making could be a capability of avionics systems,” he says. “For example, AI-enabled flight management systems (FMS) can suggest courses of action for the crew to reduce workload for both civilian and military aircraft.”

The benefits extend beyond only flight operations to handle maintenance and system management as well.

“AI-enabled avionics systems can reduce the need for interaction for some aspects by automatically handling tracking system performance failures and anticipating future maintenance needs,” Shah explains. “This will lead to enhanced safety and better crew performance and allow for greater focus on critical flight functions.”

The role of open architecture

New avionics, like all military technology upgrades, are mandated by the U.S. Department of Defense (DoD) to leverage a modular open systems approach (MOSA). Open architectures in avionics systems like the Future Airborne Capbility Environment, or FACE, Technical Standard enable faster upgrades and easier integration across platforms through reuse of software APIs.

There’s a push by the government and by industry to embrace the concepts that are more aligned with an open systems architecture… [with a] transition away from traditional integrators and their solutions,” Walters says.

MOSA is ultimately about speed and adaptability, he adds. “They want to have more agility on the front lines and have the ability to iterate different technologies more rapidly. There’s a desire to experiment with different solutions that are less traditional. All signs point toward wanting to do things much faster, have more agility, and [have] the ability to rapidly onboard and offboard different capabilities based on what the mission is – being able to plug and play.”

The Collins spokesperson emphasizes the economic benefits of this design philosophy. “Open-architecture design is foundational to the evolution of avionics. It plays a pivotal role in allowing for faster and more affordable integration on existing and future platforms – promoting interoperability, flexibility and scalability.”

Shah points out that standardization becomes even more critical as AI systems are integrated into avionics.

“As we move closer to the integration of AI into avionics systems, open standards and architecture need to evolve to the data and models that enable AI functionality,” he says. “Without a standard framework, having a variety of AI model standards and processing approaches may challenge the adoption of AI across a wide variety of aircraft platforms with consistency.”

Adapting to cybersecurity, EW threats

As aircraft become more connected, cybersecurity has evolved from an afterthought into a fundamental design consideration in avionics systems, with manufacturers implementing increasingly sophisticated protections against both conventional and emerging threats.

Walters points to recent conflicts as evidence of the growing cyber battlefield. “One of the biggest challenges today, as we’ve seen in contemporary conflicts, is that everyone has a mobile device no matter if it’s on the good or bad side,” he says. “Cybersecurity is important because you can’t have an adversary penetrate your network. We’ve seen where adversaries have taken friendly unmanned systems just based on open source reporting out there.”

Shah explains that avionics designers are implementing multiple layers of security protocols.

“The use of zero-trust architecture (ZTA) models allows for continuous authentication, end-to-end encrypted communication, and secure boot and firmware authentication, ensuring no malware can enter the system,” he says.

Shah says that AI technologies are also being leveraged to enhance cybersecurity. “AI-driven machine learning monitoring can detect potential cyber threats, helping to prevent more aggressive cyberattacks on avionics systems.” (Figure 3.)

[Figure 3 | Aitech’s A230 Vortex is a rugged edge computer designed for AI processing and signal analysis in harsh environments.]

Countering electronic warfare (EW) threats is also part of securing the flight avionics, requiring protective measures built into the core design.

“Electronic warfare resilience begins in the initial systems design stage,” Shah explains. “Anti-jamming or anti-spoofing capabilities protect navigation systems from disruption, while the use of software-defined radio (SDR) allows for flexibility through dynamic frequency-hopping, helping to avoid interference in RF-contested environments.”

Looking ahead, Shah warns that current encryption methods may not be sufficient against future threats.

“Cyber protection will also need to take into consideration the challenges associated with quantum computing, which has the potential to break traditional encryption methods,” he says.