Using GenAI to achieve operational excellence

Paul Adams is a Director and Aerospace & Defence sector specialist at Vendigital. He recently shared his insights with Aerospace Manufacturing.

The boom in Generative AI (GenAI) is creating new opportunities for aerospace and defence sector manufacturers to add value and achieve operational excellence. For some, this could seem like buzzwords or technology of the future, but it is already making a practical impact today. Used correctly, GenAI capability can help to optimise production processes, improve supply chain resilience, ringfence knowledge and reduce costs.

Many manufacturers in high-value sectors such as aerospace and defence have amassed vast quantities of data – both proprietary data about their own processes and functions, and publicly available datasets from drones and satellites, for example. However, creating tools that can sift through all this information and deliver meaningful insights to decision-makers quickly has proved challenging. For many organisations, the data opportunity has not yet been realised.

In recent years, however, the rise of Large Language Models (LLMs) has led to the emergence of new GenAI tools with much greater computational power. Using applications such as ChatGPT and Bard, end users have built bespoke models to create novel outputs in the form of text, video, audio, simulations or code. For manufacturers in the sector, it is clear that GenAI models are capable of so much more than their traditional counterparts, and new use cases are emerging all the time.

The International Air Transport Association (IATA) has recently published a report for the aviation industry, which calls for a collaborative approach to support companies in leveraging the benefits of AI and its generative capabilities. At its Innovation Day 2023, industry leaders came together to share examples of use cases for GenAI and their early experiences.

For manufacturers, a primary area for the application of GenAI models is optimising production processes from design stage through to the factory floor and delivery. For example, generative design algorithms have been developed to help the aerospace industry combat the effects of climate change, by developing aircraft that produce fewer emissions.

GenAI-enabled design allows engineers to understand and consider the cost and carbon implications of switching to green technologies, such as battery-powered electric aircraft or green hydrogen fuel. These advanced models also provide a better understanding of what needs to change for each technology to become viable at scale. For example, for electric aircraft to become more viable, batteries with a better power-to-weight ratio are needed.

NASA has developed GenAI-designed aerospace hardware, which is stronger than its human-designed counterparts and a third lighter too. Known as Evolved Structures, the generative design process has been used to create metal components, including brackets and mounts, for space missions. NASA has also recently launched a geospatial foundation model that uses satellite data to perform geospatial analysis more quickly than traditional methods.

As well as hardware, GenAI can be used to produce software too. AI generation tools that produce code based on natural language inputs can develop bespoke solutions much more quickly than would be possible otherwise. Despite being a relatively new area of application, text-to-code generation has enormous potential for manufacturers; helping them to streamline processes and improve operational efficiency.

Another key use for GenAI models is supply chain management, particularly when used for demand planning. Despite the long-term nature of most aerospace programs, the complexity of the products made by many OEMs and Tier 1 manufacturers means that demand instability can cause major issues which become amplified further down the supply chain. AI-assisted demand management models could therefore act as a ‘copilot’, alongside a human manager, and when in place they are proven to be around 25% more accurate than unassisted planning. With working capital continuing to grow this level of improvement in demand accuracy would have a profound impact on cash flow throughout the sector.

The global shortage of skilled labour is a major productivity issue for many aerospace and defence companies. As a result, many are looking to improve knowledge management within their organisations to ensure valuable capability and key data based on past interactions with suppliers and customers is captured and shared.

By applying a Chat GPT-style model, large organisations can retrieve and use knowledge that has accumulated over time in its existing systems. This might include historical information about supplier negotiations or contractual terms agreed with a large customer. By integrating this with existing tools and systems, these insights could be served up at the right time to help businesses make better decisions more quickly. In a defence context, this use of GenAI needs to be carefully planned and controlled, but trials in secure closed networks have already been carried out. The use of GenAI to support training initiatives is also proving beneficial in the defence sector, as it can be used to enhance simulation environments and create realistic scenarios for training purposes.

GenAI is a game changer for the aerospace and defence sector. Whether its optimising supply chain management processes, speeding up the design of new products or finding novel ways to get the best out of employees, manufacturers that prioritise investment in this area will be well on the way to optimising efficiency and achieving operational excellence.