Advancing Probabilistic Machine Learning to Deliver Safer, More Efficient, and Predictable Air Traffic Control
Lead Research Organisation:
The Alan Turing Institute
Department Name: Research
Abstract
The ambition of this partnership between NATS and The Alan Turing Institute is to develop the fundamental science to deliver the world's first AI system to control a section of airspace in live trials.
Our research will take a hierarchical approach to air traffic control (ATC) by developing a digital twin alongside a multi-agent machine-learning control system for UK airspace. Furthermore, the partnership will develop technical approaches to deploy trustworthy AI systems, considering how safety, explainability and ethics are embedded within our methods, so that we can deliver new tools which work in harmony with human air traffic controllers in a safety-critical environment.
Little has changed in the fundamental infrastructure of UK airspace in the past 50 years, but demand for aviation has increased a hundredfold. Aviation 2050, a recent government green paper, underlines the importance of the aviation network to the prosperity of the UK to the value of £22 billion annually. Yet our nation is at risk without rapid action to modernise our airspace and control methods, to ensure they can handle a future increase in UK passenger traffic of over 50% by 2050 and new challenges arising from unmanned aircraft, both against a backdrop of increasing global pressures to transform the sector's environmental impact.
The augmentation of live air traffic control through the use of AI agents which can handle the complexity and uncertainties in the system has transformative potential for NATS's business. This will positively impact live operations, as well as a research tool and training facility for new ATCOs. Correspondingly, NATS's research vision is to exploit new approaches to AI that enable increases in safety, capacity and environmental sustainability while streamlining air traffic controller training.
The anticipated benefits of AI systems to air traffic control have come at a critical time, providing us with an opportunity to respond effectively to the unprecedented challenges which arise from a triad of crises: the coronavirus 2019 (Covid-19) pandemic, Brexit and global warming. The UK must develop independent technical advances in the sector, without compromising sustainability targets.
The Alan Turing Institute is positioned at the rapidly evolving frontiers of probabilistic machine learning, safe and trustworthy AI and reproducible software engineering. Matching this with the world-leading expertise of NATS, supported by a world-first data set of more than 20 million flight records, means that this partnership is in a unique position to build the first multi AI agents system to deliver tactical control of UK airspace.
Our research will take a hierarchical approach to air traffic control (ATC) by developing a digital twin alongside a multi-agent machine-learning control system for UK airspace. Furthermore, the partnership will develop technical approaches to deploy trustworthy AI systems, considering how safety, explainability and ethics are embedded within our methods, so that we can deliver new tools which work in harmony with human air traffic controllers in a safety-critical environment.
Little has changed in the fundamental infrastructure of UK airspace in the past 50 years, but demand for aviation has increased a hundredfold. Aviation 2050, a recent government green paper, underlines the importance of the aviation network to the prosperity of the UK to the value of £22 billion annually. Yet our nation is at risk without rapid action to modernise our airspace and control methods, to ensure they can handle a future increase in UK passenger traffic of over 50% by 2050 and new challenges arising from unmanned aircraft, both against a backdrop of increasing global pressures to transform the sector's environmental impact.
The augmentation of live air traffic control through the use of AI agents which can handle the complexity and uncertainties in the system has transformative potential for NATS's business. This will positively impact live operations, as well as a research tool and training facility for new ATCOs. Correspondingly, NATS's research vision is to exploit new approaches to AI that enable increases in safety, capacity and environmental sustainability while streamlining air traffic controller training.
The anticipated benefits of AI systems to air traffic control have come at a critical time, providing us with an opportunity to respond effectively to the unprecedented challenges which arise from a triad of crises: the coronavirus 2019 (Covid-19) pandemic, Brexit and global warming. The UK must develop independent technical advances in the sector, without compromising sustainability targets.
The Alan Turing Institute is positioned at the rapidly evolving frontiers of probabilistic machine learning, safe and trustworthy AI and reproducible software engineering. Matching this with the world-leading expertise of NATS, supported by a world-first data set of more than 20 million flight records, means that this partnership is in a unique position to build the first multi AI agents system to deliver tactical control of UK airspace.
Publications
Pepper N
(2023)
A probabilistic model for aircraft in climb using monotonic functional Gaussian process emulators
in Proceedings of the Royal Society A: Mathematical, Physical and Engineering Sciences
| Title | Developed and implemented ATC metrics for Machine Learning. |
| Description | Developed and implemented ATC metrics for Machine Learning. |
| Type Of Material | Computer model/algorithm |
| Year Produced | 2022 |
| Provided To Others? | No |
| Impact | Developed and implemented ATC metrics for Machine Learning. |
| Title | Extensive, research ready ATC dataset with unified access to aircraft flight plan data, aircraft instruction data, aircraft coordination data, real world airspace data, real world radar data, aircraft communications data. |
| Description | Extensive, research ready ATC dataset with unified access to aircraft flight plan data, aircraft instruction data, aircraft coordination data, real world airspace data, real world radar data, aircraft communications data. |
| Type Of Material | Database/Collection of data |
| Year Produced | 2022 |
| Provided To Others? | No |
| Impact | Extensive, research ready ATC dataset with unified access to aircraft flight plan data, aircraft instruction data, aircraft coordination data, real world airspace data, real world radar data, aircraft communications data. |
| Title | Processing of operational data to provide configuration of scenarios for use in Digital Twin. |
| Description | Processing of operational data to provide configuration of scenarios for use in Digital Twin. |
| Type Of Material | Data analysis technique |
| Year Produced | 2022 |
| Provided To Others? | No |
| Impact | Processing of operational data to provide configuration of scenarios for use in Digital Twin. |
| Title | Real world data sourced for optimisation scenarios using London Middle Sector (Sectors 25/26). |
| Description | Real world data sourced for optimisation scenarios using London Middle Sector (Sectors 25/26). |
| Type Of Material | Database/Collection of data |
| Year Produced | 2022 |
| Provided To Others? | No |
| Impact | Real world data sourced for optimisation scenarios using London Middle Sector (Sectors 25/26). |
| Title | Scenario generation from real-world data on London Middle Sector (LMS) working and allowing agent training on sector that will be used for validation |
| Description | Scenario generation from real-world data on London Middle Sector (LMS) working and allowing agent training on sector that will be used for validation |
| Type Of Material | Computer model/algorithm |
| Year Produced | 2022 |
| Provided To Others? | No |
| Impact | Scenario generation from real-world data on London Middle Sector (LMS) working and allowing agent training on sector that will be used for validation |
| Title | Sourced and secured data from NATS operational systems and intermediate processing for presentation and consumption. |
| Description | Sourced and secured data from NATS operational systems and intermediate processing for presentation and consumption. |
| Type Of Material | Database/Collection of data |
| Year Produced | 2022 |
| Provided To Others? | No |
| Impact | Sourced and secured data from NATS operational systems and intermediate processing for presentation and consumption. |
| Title | Behavrioural cloning software, learning fromdemonstrations by agenda and data, learing control behaviours. |
| Description | Behavrioural cloning software, learning fromdemonstrations by agenda and data, learing control behaviours. |
| Type Of Technology | Software |
| Year Produced | 2022 |
| Impact | Behavrioural cloning software, learning fromdemonstrations by agenda and data, learing control behaviours. |
| Title | Human Machine interface code which has developed the human machine interface to starling, which include the dynamic radar screen and aircraft strip display, for replay and simulation modes. |
| Description | Human Machine interface code which has developed the human machine interface to starling, which include the dynamic radar screen and aircraft strip display, for replay and simulation modes. |
| Type Of Technology | Software |
| Year Produced | 2022 |
| Impact | Human Machine interface code which has developed the human machine interface to starling, which include the dynamic radar screen and aircraft strip display, for replay and simulation modes. |
| Title | Optimisation Agent Framework for Air Traffic Control |
| Description | Optimisation Agent Framework for Air Traffic Control |
| Type Of Technology | Software |
| Year Produced | 2022 |
| Impact | Optimisation Agent Framework for Air Traffic Control |
| Title | Phoenix and Starling Simulator Frameworks & Structure - core simulation code. |
| Description | Phoenix and Starling Simulator Frameworks & Structure - core simulation code. |
| Type Of Technology | Software |
| Year Produced | 2022 |
| Impact | Phoenix and Starling Simulator Frameworks & Structure - core simulation code. |
| Title | Reinforcement Learning Framework for Air Traffic Control |
| Description | Reinforcement Learning Framework for Air Traffic Control |
| Type Of Technology | Software |
| Year Produced | 2022 |
| Impact | Reinforcement Learning Framework for Air Traffic Control |
| Title | SimpleSector & Optimisation Agents prototype code that defines the formulation of an optimisation based agent, for which a simple problem is solved. |
| Description | SimpleSector & Optimisation Agents prototype code that defines the formulation of an optimisation based agent, for which a simple problem is solved. |
| Type Of Technology | Software |
| Year Produced | 2022 |
| Impact | SimpleSector & Optimisation Agents prototype code that defines the formulation of an optimisation based agent, for which a simple problem is solved. |
| Title | Software developed for aircraft uncertainty modelling |
| Description | Software developed for aircraft uncertainty modelling |
| Type Of Technology | Software |
| Year Produced | 2022 |
| Impact | Software developed for aircraft uncertainty modelling |
| Title | Software developed for scenario generation from real-world data |
| Description | Software developed for scenario generation from real-world data |
| Type Of Technology | Software |
| Year Produced | 2022 |
| Impact | Software developed for scenario generation from real-world data |
| Title | Software developed for synthetic scenario generation |
| Description | Software developed for synthetic scenario generation |
| Type Of Technology | Software |
| Year Produced | 2022 |
| Impact | Software developed for synthetic scenario generation |
| Title | Starling - An AI Air Traffic Control Framework |
| Description | This is a the simulation framework used for the development of AI agents for Air Traffic Control. This includes a framework for simulation, AI agent training (both Reinforcement Learning Methods and Optimisation) and Also visualisation. The software is currently private, as contracts are still being work on for the prosperity partnership. The full intention is to make the software open for general research in this area. |
| Type Of Technology | Software |
| Year Produced | 2021 |
| Impact | This is providing the software interface for collaboration on the bluebird programme, and allows effective collaboration between industry driven challenges and frontier research. It provides separation of concerns, so these groups can work both together on common challenges, but independently on domain specific components. |
| Title | Synthetic airspace designed and implemented |
| Description | Synthetic airspace designed and implemented |
| Type Of Technology | Software |
| Year Produced | 2022 |
| Impact | Synthetic airspace designed and implemented |
| Description | Air Traffic Control Officer Training for AI and Data Scientists |
| Form Of Engagement Activity | Participation in an activity, workshop or similar |
| Part Of Official Scheme? | No |
| Geographic Reach | National |
| Primary Audience | Postgraduate students |
| Results and Impact | This is a knowledge transfer activity associated with ups killing AI and data science experts with domain specific knowledge around Air Traffic Control. This was a run as two independent of 1 week courses by the prosperity partnership, with real training / simulator experience given on the Air Traffic Controller course. |
| Year(s) Of Engagement Activity | 2021,2022 |
| Description | Air Traffic Control talks |
| Form Of Engagement Activity | Participation in an activity, workshop or similar |
| Part Of Official Scheme? | No |
| Geographic Reach | National |
| Primary Audience | Industry/Business |
| Results and Impact | Series of Air Traffic Control talks delivered, with operational staff at NATS leading bi-weekly scenario sessions for the project team to inform the research and code development. |
| Year(s) Of Engagement Activity | 2022,2023 |
| Description | Data Study Group challenge with Bristol University |
| Form Of Engagement Activity | Participation in an activity, workshop or similar |
| Part Of Official Scheme? | No |
| Geographic Reach | National |
| Primary Audience | Postgraduate students |
| Results and Impact | Data Study Group challenge with Bristol University- Set a research challenge for undergraduate 'study group' at Bristol University |
| Year(s) Of Engagement Activity | 2023 |
| Description | Monthly Project Management Group |
| Form Of Engagement Activity | A formal working group, expert panel or dialogue |
| Part Of Official Scheme? | No |
| Geographic Reach | National |
| Primary Audience | Industry/Business |
| Results and Impact | Monthly Project Management Group set up and running to support and advise the PIs to ensure effective delivery, efficient cross-site operation of the research, and monitoring progress towards milestones. |
| Year(s) Of Engagement Activity | 2022,2023 |
| Description | NATS seminar about Project Bluebird to invited industry attendees |
| Form Of Engagement Activity | Participation in an activity, workshop or similar |
| Part Of Official Scheme? | No |
| Geographic Reach | International |
| Primary Audience | Industry/Business |
| Results and Impact | NATS seminar about Project Bluebird to invited industry attendees. 13/01/23 Engaging with MOD, DSTL, CAA, BEIS, UKRI |
| Year(s) Of Engagement Activity | 2023 |
| Description | Presentation at AI Expo |
| Form Of Engagement Activity | A talk or presentation |
| Part Of Official Scheme? | No |
| Geographic Reach | International |
| Primary Audience | Industry/Business |
| Results and Impact | Presentation at AI Expo giving an overview of the project and its aims. Engaging with UK Gov, European Gov, Tech Industry 15/02/23 |
| Year(s) Of Engagement Activity | 2023 |
| Description | Presentation at Aerospace Research Centre UK |
| Form Of Engagement Activity | A talk or presentation |
| Part Of Official Scheme? | No |
| Geographic Reach | International |
| Primary Audience | Industry/Business |
| Results and Impact | Presentation at Aerospace Research Centre UK, made up of 11 leading UK aerospace universities. |
| Year(s) Of Engagement Activity | 2023 |
| Description | Presentation at Queen Mary university |
| Form Of Engagement Activity | A talk or presentation |
| Part Of Official Scheme? | No |
| Geographic Reach | International |
| Primary Audience | Industry/Business |
| Results and Impact | Project Co-I presented project Bluebird at Queen Mary university in Nov 2022. |
| Year(s) Of Engagement Activity | 2022 |
| Description | Presentation at The Control Room of the Future workshop |
| Form Of Engagement Activity | A talk or presentation |
| Part Of Official Scheme? | No |
| Geographic Reach | International |
| Primary Audience | Industry/Business |
| Results and Impact | Presentation at The Control Room of the Future workshop in Ghent to audience of Digital managers, data scientists and researchers in the field of digital operations and predictive analytics in safety-critical settings, utilities, transportation, control rooms |
| Year(s) Of Engagement Activity | 2022 |
| Description | Presented project Bluebird at AI Fest |
| Form Of Engagement Activity | A talk or presentation |
| Part Of Official Scheme? | No |
| Geographic Reach | International |
| Primary Audience | Industry/Business |
| Results and Impact | Two project Co-Is presented project Bluebird at AI Fest in Feb 2023. |
| Year(s) Of Engagement Activity | 2023 |
| Description | Project page published on Turing site (Project Bluebird: An AI system for air traffic control) to share information about the project as well as team and contact information. |
| Form Of Engagement Activity | Engagement focused website, blog or social media channel |
| Part Of Official Scheme? | No |
| Geographic Reach | International |
| Primary Audience | Industry/Business |
| Results and Impact | Project page published on Turing site (Project Bluebird: An AI system for air traffic control) to share information about the project as well as team and contact information. |
| Year(s) Of Engagement Activity | 2021,2022,2023 |
| Description | Quarterly Joint Steering Committee set up to provide the main strategic interface into the plan and strategy for the partnership. |
| Form Of Engagement Activity | A formal working group, expert panel or dialogue |
| Part Of Official Scheme? | No |
| Geographic Reach | National |
| Primary Audience | Industry/Business |
| Results and Impact | Quarterly Joint Steering Committee set up to provide the main strategic interface into the plan and strategy for the partnership. |
| Year(s) Of Engagement Activity | 2022,2023 |
| Description | Research engineering teams at the Turing and Exeter University held a 'Dugnad' collaborative code clean-up session to facilitate shared working. |
| Form Of Engagement Activity | Participation in an activity, workshop or similar |
| Part Of Official Scheme? | No |
| Geographic Reach | National |
| Primary Audience | Industry/Business |
| Results and Impact | Research engineering teams at the Turing and Exeter University held a 'Dugnad' collaborative code clean-up session to facilitate shared working. |
| Year(s) Of Engagement Activity | 2022 |
| Description | Two-way training activities with NATS and the Turing. |
| Form Of Engagement Activity | Participation in an activity, workshop or similar |
| Part Of Official Scheme? | No |
| Geographic Reach | National |
| Primary Audience | Industry/Business |
| Results and Impact | Two-way training activities have been run from both NATS and the Turing involving the whole project team, building skills across disciplines for both the industrial and academic team members, including an on-site, four-day familiarisation course at the NATS Corporate Centre. |
| Year(s) Of Engagement Activity | 2022,2023 |
| Description | Weather/storm avoidance subtask investigated during 3 day weather optimisation workshop at Exeter. |
| Form Of Engagement Activity | Participation in an activity, workshop or similar |
| Part Of Official Scheme? | No |
| Geographic Reach | National |
| Primary Audience | Industry/Business |
| Results and Impact | Weather/storm avoidance subtask investigated during 3 day weather optimisation workshop at Exeter. |
| Year(s) Of Engagement Activity | 2022 |
| Description | Webinar delivered to NATS stakeholders, in conjunction with NATS International Affairs on The Role of AI in Safe Efficient Aviation |
| Form Of Engagement Activity | Participation in an activity, workshop or similar |
| Part Of Official Scheme? | No |
| Geographic Reach | International |
| Primary Audience | Industry/Business |
| Results and Impact | Webinar delivered to NATS stakeholders, in conjunction with NATS International Affairs on The Role of AI in Safe Efficient Aviation- Engaging with Safety regulators, EASA, CAA, Eurocontrol, EU Parliament, Unions, other ANSPs 15/2/23 |
| Year(s) Of Engagement Activity | 2023 |