The future of air mobility

Autonomy of Future Air Mobility project to assess the impact of autonomous flight. (Credit: Image: Livinskiy/Adobe stock)

Over the past 40 years, aircraft have become increasingly reliant on automated functions to simplify tasks for aircrew and improve flight safety—but new forms of mobility such as future air taxis and fully autonomous drones herald the arrival of a new paradigm.

The Consortium for Research and Innovation in Aerospace in Quebec (CRIAQ) recently released a progress update on its Autonomy of Future Air Mobility (AMAF) project led by Thales, with partners Presagis (now CAE) and Université Laval.

The project aims to propose a structured framework for the increasing autonomy of vehicles involved in advanced air mobility—a fast-developing segment of the aerospace market that is attracting very high levels of research and development investments worldwide.

The key innovation goal of the project is to develop and test the technology needed for an air vehicle to fly safely and fully autonomously along a previously defined flight path. Project partners Thales, Université Laval and CAE are combining their expertise in digital technologies, artificial intelligence (AI), critical systems and simulation to design the components needed for the autonomous aerospace of the future.

The AMAF project’s goal is not only to revolutionize air mobility but to also create a framework that integrates new technologies safely and sustainably across industries.

Design Engineering spoke with Jean-François Gagnon, director of the Thales Research and Technology facility in Canada, about the complexity of the task ahead.

Gagnon explains that air mobility will impact the entire aerospace ecosystem, from the regulatory environment to new technology integration. Thales is leading efforts to develop not only the technical standards but also the language and operational context necessary for integrating autonomous systems into airspaces, while addressing challenges in safety and energy efficiency.

He says autonomous systems like drones will require advancements in sensor technology and machine learning to navigate airspace safely. He also notes that the aerospace industry is relatively conservative, especially when you bring machine learning and AI into an avionics solution. “In the aerospace industry, you have to think of AI innovation differently because safety is always the number one priority,” he says. The project’s primary focus is on maintaining safety of flight, more complex in the context of autonomous detection and recognition of obstacles in flight paths.

A major challenge lies in creating AI algorithms that guarantee obstacle detection even when unforeseen objects are encountered. For example, unmanned aerial vehicles (UAVs) conducting infrastructure inspections will need to avoid pylons or birds, making AI algorithms reliability crucial to maintain safety.

Certifying these technologies is another complex issue—and Gagnon stresses that it cannot be done alone. A collaborative approach, involving various partners, is essential for ensuring regulatory compliance and building trust in AI-based systems.

Beyond safety, the AMAF project is also heavily focused on environmental sustainability. He explains that optimizing energy usage within fleets of hybrid propulsion aircraft, whether electric or fuel powered, is a major challenge. In this respect, the Université Laval team’s contributions have been valuable. For instance, while fuel is more effective during climbs, electricity can be more efficient during other phases of flight. Balancing these variables across the full spectrum of operations involves intricate mathematical optimization. The project also considers how to deploy infrastructure for refuelling and recharging aircraft, striving to make air mobility greener while still meeting operational constraints and demands.

Thales is consistently focused on pursuing disruptive innovations, especially in fields like air mobility. “I would say that this is in our DNA,” Gagnon says, adding that projects like AMAF serve as experimental grounds to test new concepts and evaluate their feasibility in real-world applications.

The goal isn’t necessarily to implement all findings directly into future systems but to identify potential opportunities, risks and areas that need further development. By addressing these questions, Thales remains at the forefront of innovation, assessing both market value and technical readiness for emerging technologies. This iterative process is embedded in the company’s core innovation strategy.

TrUE AI Framework

Thales excels in designing high-end sensors such as radars, sonars and optronics, which are core to their business. Gagnon highlights that the real value of these sensors is maximized through their integration with AI, particularly in obstacle detection and vision systems. He says that what sets Thales apart is its TrUE AI framework—transparent, understandable and explainable AI. This framework is essential for ensuring trust in AI, especially for safety-critical environments. Thales aims to qualify AI systems in these domains, addressing the rigorous demands of safety and reliability.

The Thales UAS100 is a key example of how autonomous technology is advancing in air mobility, specifically within the context of Thales’ project. Gagnon emphasized the UAS100’s role as the preliminary target for developments in obstacle detection and navigation path optimization, making it a critical focus of this initiative. Designed for security, defense, and civilian operations, it offers real-time, high-resolution imaging and can operate up to 100 kilometers from its base. Equipped with advanced sensors and secure communications, the UAS100 is ideal for tasks such as border patrol, infrastructure monitoring and disaster management. Its innovative design allows for autonomous operations, ensuring minimal human intervention while maintaining safety and compliance with aviation regulations. The drone is a scalable solution for enhanced situational awareness in various environments. Thales was able to collaborate with the CED (Unmanned Aerial System Centre of Excellence in Quebec) for in-flight trials.

Challenges

A significant challenge was the limited amount of data available for training AI models, especially in comparison to the vast array of potential airborne scenarios, as the diversity of potential scenarios far exceeds the available data. “Even if we had one billion images, it’s still nothing compared to the amount of possibilities of things that we could see in the air space,” Gagnon says.

The team needed to ensure that algorithms trained on existing data could accurately identify new, unseen objects, such as pylons or other airborne hazards. This critical technical hurdle requires advanced algorithm development to ensure safety and reliability in diverse conditions.

The collaboration with CAE is helping to mitigate this issue. CAE was tasked with creating a digital twin of the airspace environment to support real-world implementation of innovative perception and navigation functions developed as part of this project, and to provide training for users of these new systems. A virtual test bench will be created to analyse interactions between the real world and an entirely simulated environment.

Beyond this challenge, the project is always evolving. Since the beginning, project hypotheses have evolved, refining their understanding of potential system users and optimizing deployment strategies, underscoring the project’s iterative and adaptive nature.

The future

While the technological capabilities for autonomous flight may be within reach sooner than expected, the challenges extend beyond just innovation. The regulatory landscape, integration with legacy systems and industry-wide coordination will need time to adapt.

In the near future, the possibilities for autonomous flight seem promising, particularly in enhancing safety and efficiency. Regardless of the type of aircraft—whether business jets or air taxis—the integration of autonomous systems will likely progress. Collaborative autonomy, where AI works alongside human pilots, will improve safety, especially in situations where human factors such as fatigue or incapacitation come into play. “I see Canada is well positioned to contribute to this,” Gagnon says. “We have a very strong aerospace industry. We see investment in this area every year.”

Aircraft with the ability to operate autonomously in critical moments will help make air travel more secure, potentially reducing accidents caused by human error.

While fully autonomous flight may not become mainstream anytime soon, projects like AMAF are helping to lay the groundwork. “It’s not going to happen within the next five years,” Gagnon says. “But it’s time to work on it nevertheless, because it will come.”

He also emphasized the long-term impact of the project in cultivating the next generation of engineers given the number of people they’re training on these emerging topics. “We did one thing I’m always proud of these kinds of projects: We’re building the future of Canada for engineers that will work on these real things in the future. In the next 20 years, they’ll be leading the show. So that’s the thing that’s most important for me.”

• Tweet

• PAMI brings new life to machinery with custom parts, modern controls
• Designing Sanofi’s new vaccine facility in Toronto