As a member of the AIAA AVIATION Forum 2026 Guiding Coalition, it was clear to Tracy Elving that a nuanced discussion about artificial intelligence had to kick off the forum. When asked about the core value of AI, some on the Guiding Coalition felt that it could provide solutions to ambitious technical challenges, while others argued it could enable engineers to do higher order tasks. A lack of consensus created a provocative opportunity.
With over 40 years of experience as an aeronautical systems engineer and current Senior Director of Corporate Relations for Aerospace Engineering at the University of Illinois Urbana-Champaign, Elving has a unique perspective on the tools and training composing current discussions about the role of artificial intelligence in technical aircraft design.
If speed is what we want, does AI actually provide a solution or does it create new problems?
She believes that we have reached an inflection point between valuing productivity and performance. While current aircraft in production have largely maximized efficiency and performance gains, there is a current multi-year production backlog. At the same time, major space industry disruptors claim to be able to bring payload launch costs below four figures.
“The only way that we solve these things is through speed. Productivity makes the market. We have 40 years’ worth of new tools, but we don’t do anything faster because we have focused on complexity.”
Elving determined that a bifurcated approach to programming would be the best way to broach the challenge, as the focus is not solely AI, but on what AI does in relation to the other tools and methods which precede it.
Without all these scenarios modeled does AI have the data necessary to help identify all the risks?
On Monday, 8 June, Graham Warwick, Executive Editor of Technology for Aviation Week, will moderate “What Problems are We Really Trying to Solve?” and ask public and private sector leaders about how technology assists in current approaches to safety, testing, and resolving market realities for innovation and disruption. The discussion will aid in contextualizing AI as an assistive tool and prospective challenge.
In Elving’s Systems Engineering course at the University of Illinois, the focus is notably light on equations and heavy on decision-making and risk-taking. Students need to gain this experience and understand that if they are going to use a tool to support decision-making, they need to know what problem they are trying to solve.
“Safety is paramount in our industry. To achieve the highest safety levels we don’t trust just one single system and build in the necessary redundancies to ensure those levels. When there is a catastrophic failure, it is typically due to multiple system failures occurring at one time that weren’t anticipated.”
Do current perceptions of the speed of AI obscure its reliability?
On 9 June, Elving will moderate, “AI Hype vs. Reality – Past vs. Prediction.” In this Forum 360 session, she aims to explore potential uses and concerns for AI in model-based testing, efficiency, and decision-making. While speed contributes meaningfully to market viability, in recent years complexity has been prized and the two elements may be at odds. Speed requires some level of simplicity and replicability whereas aircraft design complexity follows the laws of physics. However, design and production systems are multidimensional.
“This is not about being anti-tool but deploying the right tool in the most effective manner to solve the problem at hand. – You don’t pull out a wrench to hammer in a nail.”
She noted that robots can be programmed one way and will make a widget the way it is told. If there is a mechanical error or incorrect programming in production, the system may stop, or you may have 300 defective items. However, people have the ability to make decisions, de-risk based on novel scenarios, and they will work until the widget is perfect.
“We need to set a clear goal and figure out how to solve it. ‘Invention’ is a broad response to a problem statement. The end goal is not AI as we don’t produce AI – we produce aircraft. We need to figure out how to use AI and for what purpose to achieve our objectives.”
AIAA AVIATION Forum 2026 runs 8–12 June at the Manchester Grand Hyatt, San Diego, California. Registration is available at aviation.aiaa.org.
