When AI Empowers Aerospace

In May 2023, I had the opportunity to be invited as a speaker at the AED Days 2023 in Oeiras, near Lisbon, Portugal, a major gathering of various stakeholders in the aerospace industry in Europe. Industry leaders were present, which allowed me to engage with companies such as Airbus, Aernnova, Embraer, Lockheed Martin on topics related to digitalization and the contribution of AI in digital transformation.

With Quebec and Canada enjoying a highly esteemed reputation in the field of artificial intelligence in Europe, I was proud to represent Canada at this significant gathering, alongside the representatives of the Canadian Embassy. In this article, I am delighted to highlight the key highlights from the various discussions I had with domain specialists. Needless to say, AI was at the center of the conversations!

Mature Engineering Expertise... But Challenges for Transitioning to AI

One might think that aerospace companies, producing highly sophisticated aircraft, are the most mature when it comes to data utilization and the use of artificial intelligence. After all, these companies produce or contribute to the development of highly sophisticated aircraft that travel through the air and even into space! It is true that they have extraordinary engineering expertise, often spanning several decades. The quality of the aircraft they build is crucial, and these companies have great concern for it. Their industry maturity is high, and the reputation of the players is well-established.

However, when it comes to data and digital transformation, aerospace companies face similar challenges to many large organizations in other sectors that are not digitally native. An amalgamation of previous acquisitions, autonomous business units, multiple and diverse generational technology systems has led to data silos and real challenges in data integration and matching. Moreover, operational processes that have been in place for many years, if not decades, mean that companies still have to deal with information stored in paper documents. And naturally, the quality of data can vary greatly, as several processes are still manual. Incomplete or non-compliant data issues are encountered. On the other hand, the modernization of various manufacturing equipment is generating an increasing amount of data to be stored, managed, and leveraged. Significant human investments are therefore required in terms of data expertise within organizations.

Several characteristics of the aerospace industry also contribute to the slow and cautious development and deployment of AI. One particular aspect I would like to mention is that, unlike the automotive industry, production volumes in aerospace are relatively low. We don't manufacture millions of airplanes every year! Instead, the focus is often on specific development, sometimes even approaching customized solutions. This aspect needs to be considered in the context of intelligent process automation. The potential impacts relate to the volume of data, the applicable AI approaches, as well as the underlying financial model for AI solution development and deployment. Another crucial factor to consider is the risk associated with making changes to processes. When it comes to critical processes involved in aircraft assembly or space-bound vehicles, experimentation (and especially errors) is not desirable. Ensuring the safety of individuals must always be a top priority. Therefore, complete trust in the deployed AI solution is essential when it comes to critical processes!


The Potential of AI in the Aerospace Industry

For the manufacturing and industrial sector, at Videns, we offer a straightforward model that assesses the impact of data valorization and AI on company activities. This model consists of three categories:
1. End-to-end internal digital transformation;
2. Digital transformation of products and services;
3. New digital business models. This model is illustrated in the following figure:

In my opinion, this model is highly relevant when discussing the impact of data valorization and AI in manufacturing and industrial organizations in the aerospace sector. The first way to make an impact is by focusing on the transformation of internal operations within the company. This primarily involves improving operational efficiency or effectiveness through the use of AI. The return on investment is often substantial and helps address challenges such as labor scarcity, which is also a concern in the European and Canadian aerospace industry. Projects in this category often aim to reduce costs. Before implementing intelligent automation with AI, it is essential to optimize the process beforehand. We don't want to apply AI to something that is inefficient or suboptimal!

The second way to make an impact with AI is to integrate this tool at the core of the services or products being distributed. This involves modernizing the offering through AI. The benefits are numerous: gaining a competitive advantage by providing a more advanced product or service than competitors, improving customer satisfaction, reducing human errors during use, simplification, and more.

Finally, AI can be leveraged to generate new digital business models or new products and services. This often comes from the contribution of numerous startups to the aerospace industry! It involves thinking outside the box or discovering unmet needs that can be addressed by AI.


Some Use Cases

Companies are often highly optimized in their core business. It's their expertise and differentiating factor. Use cases in functions outside of the core business, where processes are often less optimized, present great opportunities. For example, a company manufacturing aircraft engines may have multiple suppliers. However, it's highly possible that the supply chains, which are not at the core of the company's activity, may not be optimized to the same level as the engine manufacturing itself.

At the AED Days, the discussed use cases revolved a lot around predictive maintenance and automation of production chains. Predictive (and proactive!) maintenance enables the detection of anomalies in production equipment and predicts potential breakdowns. This allows for anticipating the need for maintenance and planning it before problems occur. Such solutions help minimize downtime and, in the case of aircraft, reduce the costs associated with non-operation that occur during shutdowns.

Automation of production chains through AI is closely related to equipment modernization. New equipment often integrates automation that enables simple tasks to be performed automatically. With the data generated by sensors and modern installed devices, more advanced automation solutions can be developed and integrated.

In addition to the discussed use cases, I would add that AI and optimization approaches can also support processes such as cabin layout. The available materials, their volume, cost, production times, and project schedules are examples of constraints that need to be considered to produce with improved efficiency and profitability.

Lastly, predictive solutions can have a significant positive impact in the aerospace sector. Predictive models can be developed to anticipate parts demand, material costs, or, for airlines, airplane ticket purchases. Such AI solutions enable better cost and time management, improved resource allocation, and dynamic pricing, among other benefits.

AI for the Benefit of the Environment and Sustainable Development?

The aerospace industry is fascinating! During the AED Days, I had the opportunity to attend very inspiring lectures that took us on a journey into space. Seeing the industry's potential beyond Earth makes us feel very small... However, real concerns were also discussed, particularly regarding the aerospace industry's impact on the environment and the pollution generated in space by the multiple debris resulting from human activity. During his lecture at the AED Days, Raphael Roettgen from the organization e2mc.space warned that the human impact on the space environment around Earth is comparable to the devastating issues of plastic waste currently present in the oceans. It made a strong impression on me. In 30 years, will we be trying to find solutions to clean up space? Could the human impact on the newly explored environments be more positive?

A challenge is therefore set: how can AI support the aerospace industry in terms of sustainable development?