Skip to main content
search
AIIndustry Insider

AI in the Skies: Generative AI for Aerospace & Defense

By August 1, 2023September 9th, 2023No Comments

As much as the idea of artificial intelligence (AI) exploded into public consciousness in 2023, it’s not an entirely new concept. What we’ve seen is an explosion of “generative AI,” an AI system that can take inputs and create something entirely new, as opposed to traditional AI, which can make decisions based on human inputs. These days, we’re so used to tools such as automated assistants (Siri or Alexa), predictive text, or even computers playing chess, that we hardly register those as AI. And just like the rest of us, manufacturers are also starting to transition from traditional AI models to generative.

Aerospace & Defense (A&D) manufacturers have been leveraging AI for some time, most obviously for autonomous aircraft (UAVs or drones), but also for sensor and feedback systems that monitor data to enable product operations. As Aviation International News (AIN) puts it, in their article “Beyond Automation: How AI Is Transforming Aviation,” “The aviation industry already has used at least some primitive form of AI technology for years, particularly for manufacturing and MRO [maintenance, repair, and overhaul processes].” 1

New AI models and capabilities are allowing A&D manufacturers to push the envelope beyond merely analyzing data and making predictions, using it instead to create new data similar to its training data. In the case of A&D, that means using AI to perform more routine operations, analyze more data faster, and ultimately contribute to fleet optimization, flight planning, and ground operations.

AI and the Digital Twin

Most A&D use of AI goes hand-in-hand with the use of digital twins—digital representations of a physical product, system, or process. The digital twin is the basis for the application of real-world data for simulation, integration, testing, monitoring, and maintenance, all without having to create physical prototypes and perform physical testing. Digital twins have proven hugely valuable across industries already, even with the “weaker,” traditional AI systems currently in play. But generative AI promises to be even more useful, not only to utilize digital twins, but also to build them.

As AIN notes, “While digital twins can help save time and resources, they’re also expensive and time-consuming to create. But generative AI will soon make the process of building digital twins much faster … for just a small fraction of what it costs today.” With greatly reduced cost—one expert estimates future digital twins requiring only as little as 1% to 10% of current cost—A&D manufacturers will be able to make more widespread use of digital twins, perform more analysis, and optimize product performance and operations quicker and across more of the organization.

AI and the Human Partner

One of the concerns people have about AI and industry—as is often true for any new technology or automation—is that AI will take over, doing away with traditional roles filled by humans. However, experts see roles changing, not being eliminated, and they emphasize the complementary relationship between machines and human staff.

Sometimes AI has the advantage, and sometimes humans do. AIN quotes Rishi Rajan, founder and CEO of GridRaster, a software company that employs AI and spatial mapping software, saying, “’The human in the loop is always going to be there,’ because AI, while good at pattern recognition and making predictions, will never improve on human perception.”

At the same time, AI is better and faster at processing, analyzing, and drawing conclusions from huge volumes of data than humans—though so far, AI only works if you have enough good data to train the system with. In “The Coming Age of Artificial Intelligence,” from Military & Aerospace Electronics, Bryan Nousain, of the U.S. Naval Research Laboratory, discusses how AI can help fill gaps. “AI and machine learning approaches, such as deep neural networks, fill the gap when physics-based models are not available. However, there may not be adequate data for training a deep neural network prior to its deployment. This is driving the development of generative models to augment training data for supervised and reinforcement learning applications.” 2

AI and the Future

Experts have different ideas for future AI development. Rajan, in the AIN piece, predicts “the greatest value of generative AI models like ChatGPT will come when aerospace companies begin to verticalize the technology, integrating it with their own intellectual property for internal use.” This would open “a world of new use cases for AI” across the greater manufacturing industry.

On the more purely Defense side, Edge AI seems to be the newest buzzword. The idea of embedding AI into or onto less-sophisticated systems—whether that’s satellites, drones, or even personnel carriers—means generating even greater volumes of data that can be analyzed to improve products, improve processes, and save human lives.

As yet, those are predictions of where AI might develop. For the moment, AI is still in the early stages of adoption throughout the Aerospace & Defense industry, as manufacturers understand its capabilities and limitations and determine how to utilize it fully. But the benefits are already being realized.

1 Hanneke Weitering; June 14, 2023; “Beyond Automation: How AI Is Transforming Aviation”; www.ainonline.com; https://www.ainonline.com/aviation-news/aerospace/2023-06-14/beyond-automation-how-ai-transforming-aviation

2 Jim Romeo; April 19, 2023; “The coming of age of artificial intelligence”; www.militaryaerospace.com; https://www.militaryaerospace.com/computers/article/14290944/artificial-intelligence-machine-learning

Close Menu