Cooperation between AI and humans, data generation, and algorithmic understanding are some of the themes of the future of aviation. Through four European research projects, Deep Blue is consolidating its leadership in this field.
Deep Blue , the leading SME in Italy in terms of number of research and innovation projects funded under the European Horizon programmes (source: European Commission ), with over 20 years of experience in the aviation sector, has been working for over a decade on the inclusion and enhancement of Artificial Intelligence in this sector.
Stefano Bonelli , Head of Innovative Human Factors , commented on Deep Blue’s experience in this field: “Deep Blue began working on artificial intelligence when machine learning algorithms proved useful for obtaining real-time insights into the workload of air traffic controllers and pilots. We immediately used innovation to measure human performance and develop hybrid teams in which humans and digital assistants share tasks and responsibilities . Even today, as Human Factors experts, we are focused on finding ways to make these teams efficient and safe.”
FOUR EUROPEAN PROJECTS TO LOOK TO THE FUTURE
Today, artificial intelligence can assist in many aspects of aviation. It can help improve safety by identifying potential risks in real time by analyzing large amounts of data from sensors, radar, and other systems, with a level of analysis and speed superior to currently used tools. It can also help predict and prevent accidents through warning systems. It can be used for predictive maintenance , predicting possible aircraft failures or problems. Furthermore, algorithms can analyze weather data , air traffic, and other variables to optimize flight routes and reduce travel times and fuel consumption.
Artificial Intelligence can support pilots and ground crews in analyzing data and making complex decisions , for example during emergency situations or in optimizing flight operations. In training and simulation, AI can be used to train pilots, air traffic controllers, and maintenance personnel in realistic and complex scenarios, improving the effectiveness of training. Last but not least, AI can automate processes such as flight planning and load management and support controllers in air traffic control.
Deep Blue is addressing these issues through four European research projects that aim to make significant progress.
An algorithm for air traffic management
ASTRA is a project funded by the SESAR 3 Joint Undertaking, involving a multidisciplinary team of five partners from four different countries: Malta, Spain, Italy, and Switzerland. The consortium includes academic institutions, air navigation service providers, software developers, and, of course, Deep Blue. The partners have complementary expertise, combining academics, technicians, researchers, and human factors experts.
By developing an innovative tool based on Artificial Intelligence , ASTRA aims to transform the way air traffic is managed and mitigate congestion hotspots .
Deep Blue is responsible for the research and experimental design, as well as coordinating the dissemination and communication actions .
The AI-based predictive tools developed by ASTRA will transform the way traffic flows are managed, supporting flow managers by increasing their ability to predict issues (e.g., too many aircraft scheduled for a given airspace sector), improving traffic planning, and creating a more efficient and sustainable aviation ecosystem.
Artificial Intelligence to generate synthetic data
SynthAIr (Synthetic Data Generation for ATM Systems) is a project funded by SESAR and Horizon Europe and is a milestone in starting to define a new path for the use of AI-based tools for synthetic data generation in aviation.
With the rapid advancement of AI technologies, the aviation industry is at a critical juncture where data availability poses significant challenges to the development of new AI-based tools, especially those related to safety-critical data .
As is well known, training artificial intelligence requires a large amount of data; sometimes, this data is not sufficiently available (for example, fortunately, data relating to critical events is scarce, precisely because of the high levels of aviation safety).
SynthAIr is ready to address this pressing problem with an innovative approach.
At the heart of the project is the concept of a Universal Time Series Generator (UTG) : a model designed to learn from multiple time series datasets and generate synthetic data that accurately represents new, previously unseen datasets . In the aviation industry, for example, the UTG can be trained on data from specific airports and then used to generate synthetic data for entirely new airport environments.
Within this Deep Blue project, he is working on evaluating the effects of the developed methods and forecasting models on the performance of air transport operations, as well as coordinating communication efforts among all European partners.
Remote Digital Tower: The Industry’s Latest Innovation
The TRUSTY project , funded by the SESAR 3 Joint Undertaking, aims to implement the use of Artificial Intelligence in Remote Digital Towers (RDTs).
TRUSTY works in the field of air traffic management to promote the ever-increasing use of artificial intelligence, focusing on improving transparency and explainability (ensuring that the AI’s behavior and proposed solutions are always clear and understandable to operators) and improving trust in new tools.
RDT offers the ability to manage airport traffic flows remotely: while conventional control towers house operators with direct visual supervision of runways and taxiways, RDT uses video transmission to provide the exact same information. This advancement enables the provision of airport air traffic services from virtually anywhere, promising significant improvements in operational efficiency, safety, and increased controller situational awareness.
The project incorporates information visualization techniques such as visual analytics , data-driven storytelling , and immersive analytics into human-machine interactions.
The ambition in this context is to enable workers to preserve and maximize their cognitive resources to best utilize them in essential tasks. In this regard, having transparent, reliable, and explainable algorithms can make a significant difference in their implementation. Deep Blue will be responsible for testing the algorithms developed by TRUSTY.
Deep Blue will analyze the results achieved through the drafting of the “Validation Report” to draw key conclusions on the operational feasibility and benefits of the proposed solutions. It will also develop guidelines for assessing the reliability of AI-based decisions in the context of remote digital towers.
Development of hybrid human-machine teams
CODA (COntroller adaptive Digital Assistant) is a project that aims to increase the efficiency, capacity, and safety of air traffic management by enhancing collaboration between humans and artificial intelligence and developing a system in which tasks are performed collaboratively by hybrid teams, but also dynamically assigned according to the principles of adaptive automation.
Key aspects of the project include monitoring and predicting the mental states and tasks of air traffic controllers, as well as defining strategies to keep operators in optimal condition in terms of workload, fatigue, situational awareness, and stress. At the same time, the aim is to improve the overall efficiency of the joint human-automation work system.
CODA will develop a prototype system that can anticipate and prevent human performance issues to dynamically guide task assignments between air traffic controllers and digital assistants.
The project is coordinated by Deep Blue, which is responsible for human-AI teaming, prototype validation, and final results generation.
Stefano Bonelli concluded: “ In aviation, the main priority is safety: the level of attention to possible risks and therefore the number of checks that must be passed to be authorized to use new technologies is very high, and this is one of the reasons why aviation is one of the sectors with the best safety record.
The introduction of AI in aviation is being driven at the European level, and there is collaboration between various safety and certification bodies, industries, and research.
For example, EASA (the European Aviation Safety Agency) proposes several steps: Human assistance/augmentation (2023-2025+), human-AI teaming (2025-2035+), advanced automation and autonomous AI (2035-2050+), in which the level of automation and autonomy is gradually increased.
Broadly speaking, in the field of research, AI ranges from supporting operators (such as controllers) to managing complex traffic flows across Europe, from detecting fallen objects on airport runways to developing new materials and performing preventative maintenance. In short, the opportunities are numerous, and there’s a lot of work and excitement surrounding each of them. AI can lead to greater operational efficiency, improvements in safety and reliability, as well as a better overall experience for passengers and operators in the aviation industry.
The company: Deep Blue srl
Deep Blue is an Italian SME based in Rome that addresses today’s most important social and technological challenges through advanced research and consulting. Since 2001, the company has contributed to the development of research in Europe, securing EU funding for over 110 projects, including 14 as coordinator. It is the leading Italian SME and the third largest in Europe for research and innovation projects awarded and implemented within the Horizon framework program (from 2014 to present, source: European Commission ).
SESAR3 Joint Undertaking
SESAR 3 is an institutionalized European public-private partnership established to accelerate the implementation of Europe’s digital sky through research and innovation. The partnership is developing cutting-edge technological solutions to manage conventional aircraft, drones, air taxis, and vehicles flying at higher altitudes. The SESAR 3 JU partnership brings together the EU, Eurocontrol, and more than 50 organizations spanning the entire aviation value chain, from airports, airspace users of all categories, air navigation service providers, drone operators and service providers, the manufacturing industry, and the scientific community. The partnership also works closely with regulatory and standardization bodies, particularly EASA and Eurocae, as well as key stakeholders such as professional staff organizations, the space and military communities, and global partners.
| Contacts:
Deep Blue Press Office: Silvia Magna +39 349 25 16221 |
For further information:
Human-AI Collaboration: First Guidelines from the Air Force Air Traffic Flow Management: AI Avoids Traffic Jams in the Skies |