Transforming Flight Safety and Performance: Adaptive AI Flight Control Systems for Next-Generation Aviation

RedKnight is proud to announce it has supported Luffy AI Ltd and the University of Southampton with their successful application to Innovate UK’s Aerospace Technology Institute (ATI) programme, securing the collaborative partnership £792,065 in non-dilutive grant funding. The project will develop ultra-robust, adaptive AI systems for flight control, enabling safer, more resilient, and cost-effective flight operations for future generations of aircraft.

Background

The aerospace industry is constantly pushing the boundaries of innovation, with safety and performance at the forefront of all technological advancements. However, modern digital fly-by-wire control systems, which govern an aircraft’s flight operations, are typically designed around the specific aerodynamics and mass properties of each aircraft. This reliance poses significant challenges in scenarios where variations arise, such as during equipment failures or changes in the mass distribution of flexible wings. These variations can severely degrade the performance of flight control systems, requiring human intervention and potentially compromising safety. This limitation is an obstacle to the development of modular, flexible next-generation designs and the path to fully autonomous flight operations.

Project Objectives and AI Flight Control Solutions

The primary objective of the project is to address the limitations of existing flight control systems by developing adaptive neuroplastic control networks that can dynamically adjust to changing conditions and disturbances. These systems, driven by AI, will not only enhance flight safety and efficiency but will also open up new design possibilities for next-generation aviation.

Luffy AI, as the industry partner, is leveraging its expertise in developing ultra-robust AI solutions. The company is working to create adaptive AI-driven flight controllers that can handle variations in an aircraft’s aerodynamic and mass properties, allowing for real-time adjustments during flight. These control systems will be a critical enabler for the stability and control of advanced air mobility vehicles, including aircraft with modular bodies, highly flexible or morphing wings, and distributed actuation systems.

Luffy CEO, Dr Matthew Carr said, “Drones are an excellent ongoing research platform for us. The high-speed adaption to disturbances required in flight control offer an opportunity to keep improving our core AI technology that benefits our industrial customer base. Thanks to RedKnight for making this possible!”

The University of Southampton, a global leader in aerospace engineering research, brings its cutting-edge research in AI techniques for flight control to the project. In particular, the university’s domain randomisation techniques for Deep Reinforcement Learning (Deep RL) algorithms have been instrumental in ensuring the adaptability and robustness of the AI systems under various operational conditions. By using these advanced AI methodologies, the project team aims to create flight control systems that surpass current industry standards in terms of safety, performance, and cost-effectiveness.

A Stepping Stone for Future Aviation

One of the innovative approaches taken by the project is the use of low-cost drone platforms as a risk-managed and cost-effective stepping stone toward more complex civil flight operations. These drones will serve as testing grounds for both simulated and physical flight tests, enabling the team to refine and validate the AI-driven adaptive neural flight controllers in a controlled environment.

By focusing on forward-flight and vertical take-off and landing (VTOL) configurations, the project aims to demonstrate how AI-driven adaptive controllers can significantly enhance the versatility and resilience of flight systems. The use of drones as an experimental platform will allow the team to explore design possibilities for future aircraft, including those with advanced air mobility capabilities, in a safe and scalable manner.

Commercial Potential

The project has already garnered significant interest from Tier 1 channel partners, who have recognised the need for AI-driven flight control systems and are actively seeking innovative solutions to improve the safety and efficiency of next-generation aircraft. The AI features targeted by this project are aligned with early adopting use cases in logistics and heavy-lift applications, sectors that are expected to be among the first to benefit from these advanced control systems. By demonstrating the viability and performance of AI-driven flight controllers in these applications, the project aims to lay the foundation for the widespread adoption of AI in other areas of civil aviation.

Value Creation

In addition to the grant funding, the project anchors a critical technology capability within the UK. The development of AI-driven adaptive flight control systems has the potential to create significant value for the UK economy, particularly in the aerospace sector, where advancements in safety, performance, and cost-efficiency are always in high demand.

Furthermore, by supporting this project, Innovate UK is positioning the UK as a leader in AI flight control technology, with significant potential for future growth in both domestic and international markets. The innovations developed through this project could lead to new business opportunities, increased exports, and the creation of high-skill jobs in the UK’s aerospace and AI sectors.