DataBeacon is introducing the first Digital Assistant that will be ready for operational air traffic control rooms: Victor5. Victor5 is powered by an AI engine that can guide air traffic controllers through potential conflicts—all on the cloud, with no local computing capacity required. This engine runs on a flexible ML ops platform that will ultimately grow to support air traffic controllers in a variety of traffic scenarios, allowing them to safely and efficiently cope with post-Covid increases in traffic levels and potential challenges. Victor5 will be ready soon, too: Victor5 functionalities have been in development and validation since 2021 and will be validated in operational rooms in 2022.
Air Navigation Service Providers continue to face strong challenges: reducing costs, preventing delays by opening airspace capacity and maintaining safety. Environmental concerns also put pressure on providers to ensure that they monitor the environmental efficiency of routes. And while free-flight airspace allows airplanes to fly straighter routes, this increases the complexity of managing said airspace when compared to traditional route structures.
That’s why AI digital assistants will revolutionise the operational room by anticipating complex traffic scenarios, forecasting airspace sector workload and allowing more efficient and environmentally-friendly routing—all while increasing safety.
Indeed, in the coming years, AI digital assistants will make us completely rethink workload and capacity in air traffic control operations. Air Traffic Control Systems are on the brink of a groundbreaking transition—one that embraces AI and digitalisation despite the complexity of Controller Working Positions (CWP). Big data and cloud computing have paved the way for the processing of massive historical datasets, and these same datasets can now train AI algorithms to support air traffic controllers by identifying conflicts and non-conflicting traffic. This means that organizations like DataBeacon are now ready to launch tools that help provide early warning signs of complex traffic scenarios. Even more, Victor5 will be able to propose separation instructions that ensure safety whilst minimising capacity restrictions and emissions.
How will this all work? As mentioned, large historical datasets are needed to train the AI air traffic control models. Trajectory data is labelled and annotated with traffic conflicts, non-conflicting traffic, ATC instructions, meteo data and fuel consumption. Curating the training dataset and developing a big data ML platform ad-hoc for ATC is the first step to building digital assistants.
Adoption will be gradual, starting with ATC supervisors and planning positions. AI models will identify sectors that are likely to have more conflicts and, thus, require more separation instructions. A planning controller will be assisted by a portable device that receives all information from the digital twin air traffic controller, ensuring fast identification of critical conflicts and efficient resolutions that support decision-making. Built as a service, rather than as a product, these AI digital twin models will offer a perfect complement to controllers as well as strengthen contingency procedures in operations room.
DataBeacon is a start-up comprised of data engineers, AI experts, aerospace engineers and air traffic control experts, developing innovative AI tools for the aviation industry. The project Level5 was funded by the European Innovation Council in 2020 and the team has built an engineering validation prototype ready to be tested by air navigation service providers. Current functionalities can be deployed as early as 2022 and the transition period from the pandemic will be the perfect time to introduce digital assistants. Scalability in the Air Navigation Service Provision will be critical in the next years—act now and ensure you’re ready to embrace the post-Covid era of aviation!
Author: David Pérez
