Connected Places Catapult organised this competition to bring the UK’s world leading academics together with businesses to develop new and exciting research projects to prepare for the future of aviation security. The Catapult remains connected to the project by building a cohort of successful students to facilitate knowledge exchange between the project partners. The Catapult is helping to leverage £1.3m in contributions towards the projects in partnership contributions.
Future Aviation Security Solutions Industrial PHD Partnerships (FASS IPP)
FASS IPP is the Future Aviation Security Solutions Industrial PhD Partnerships programme which is bringing together academia and industry to focus on the future of aviation security.
The scheme is a jointly funded programme by the Department for Transport and the Home Office to further enhance the UK’s aviation security for the future. This will be done by funding new research in this area whilst also developing a pool of future leaders in the aviation security field.
Projects and their Industry Partner
Secure Access Control in Smart Airport
Student: Yang Chen
The primary aim of the proposed project is to enhance airport security and access control measures by leveraging the capabilities of 6G technology and machine learning to protect against potential cyber-attacks. This can be done by establishing a cutting-edge cybersecurity infrastructure utilising 6G technology to create a secure and robust network for transmitting sensitive data across the airport’s various security devices, control systems, and personnel. Moreover, the project aims to develop advanced machine learning models to analyse network traffic and access logs in real-time, identifying patterns of cyber-attacks and potential anomalies that may indicate unauthorised access attempts or malicious activities.
Overall, the project’s aim is to create a proactive, resilient, and cutting-edge airport security system that leverages 6G technology and machine learning to defend against cyber threats effectively. By doing so, the project will be a platform to enhance passenger safety, protect airport infrastructure, and maintain public confidence in the security of air travel.
Small Unmanned Aerial Vehicles (UAV) Detection and Tracking based on Stereo Vision using Machine/Deep Learning Strategies.
Student: Mariusz Wisniewski
I am detecting drones near airports with security cameras using computer vision and machine learning techniques, in particular convolutional neural networks and reinforcement learning. A large part of the project has been the creation of data – to achieve this I generated synthetic images of drones generated by 3D rendering. I use these images to train neural network models to predict whether a drone exists in an image. The advantage of this method is that it is cheaper to produce these images and accurate labels, compared with real-life images.
Securing improvements in passenger satisfaction and employee wellbeing: a behavioural study of human interactions with airport CT security technology.
Student: Gary Beaten
Robust Detection of Small Airborne Targets for Secure Society.
Student: Daniel White
Drones technology is ever improving. Their availability is increasing, and their prices are falling as newer models come out. We can all imagine the potential for harm that drones have in the wrong hands. Likewise, there are many people who want to use drones more widely for moral and civil applications. The variety of drones that exist is constantly expanding; they can be large and powerful, or small, light, and highly portable; they can be thousands of pounds or homemade. To manage increased drone usage, we are needing a way to have full total awareness of a region of sensitive airspace of drone presence. There are many options, and the more options employed simultaneously, the better, but radar allows day or night, weather resistant surveillance, and radars can be built to operate over long ranges.
So how does a surveillance radar work? It stands for radio detection and ranging – so they can measure how far away objects are – but the thing to know about modern radars is that they are able to measure the speed of an environment extremely well. Simply put, a radar sees everything in its line of sight, and discretises the world into how fast it is moving towards or away from the radar. A moving target then emerges from the otherwise stationary world and is detectable. Detection of targets is done in this way, and this is very well established within the radar world. However, a radar that is sensitive enough to detect small drones is going to be sensitive enough to detect birds also! This is problematic, as any surveillance system cannot afford to yield a high rate of false positives – birds classified by the radar as drones, which will trigger needless countermeasures and decrease operator trust in the system – or indeed false negatives – where a drone slips through undetected.
In this PhD, I have been working to understand, and ultimately improve the state of the classification challenge between birds and drones in a radar context. Large drones are much easier to differentiate from birds, as the motion of the spinning propellor blades can be measured alongside the drones body motion. The detection of these extra signals is a tell-tail sign that the target is a drone. For small drones with small blades, a radar is less likely to be able to detect the motion of the propellors – so a classifier needs to be savvy enough to leverage subtle differences between these and bird targets. Indeed, birds exist in a verity of sizes, and can fly in seemingly pattern, sometimes in groups as a flock or individually – so understanding the radar signatures of birds is another interest of my PhD which may provide some insight into how we can detect drones better.
Project name: CNN-Based threat detection in 3D CT images.
Student: Angelo Malandrakis
Innovative quantitative and quantifiable vapour generation methods for explosives detection.
Student: Nancy Whittaker
Explosive detection relies on calibration of detection equipment to a set of standards – these exist for all types of samples for explosives, however, explosive vapour samples are complicated. Current vapour generation methods that create these samples require large, expensive, time-intensive equipment and as of yet, are not remotely compatible. Current testing of explosive detection devices is also required to be taken to laboratories for calibration, whereas this project aims to reverse that – creating an in-situ explosive vapour sample generator that is field-friendly, to calibrate explosives detection equipment in airports. The technology used will be molecularly imprinted polymers – this process creates gaps complementary to the explosive we want the system to be calibrated for; it will trap a known concentration of this explosive that will then be released controllably from the polymer, possibly by heat, and connected with said vapour detection equipment. The molecularly imprinted polymers will be created on an optical fibre in order to facilitate the field-friendly approach. They will be coated with gold to induce a surface plasmon resonance effect – this induces waves that interact with the environment surrounding the gold layer, producing a wavelength shift as a result i.e. when the explosive molecules bind to the gaps. As a result of this project, vapour detection devices will be calibrated to smaller amounts of explosive vapour, becoming more sensitive and therefore, apprehending more explosive-related events.
A portable condensed-matter atomic clock.
Student: Matthew Green
Holistic Intelligent Security for Digital Aviation Connectivity and Monitoring System.
Student: Huw Whitworth
Aircraft operations heavily rely on real-time communication with airports and third-party vendors to ensure safe and reliable connectivity. Traditional communication systems are inadequate for handling large non-safety critical data volumes, while existing IP-driven systems lack the required throughput, assurance, and low latency demands. This project aims to leverage 5G technology, specifically Ultra Reliable Low Latency Communication (uRLLC) to augment existing communication topologies within aviation. By deploying a 5G aviation ecosystem, this project addresses the expansion of the airport threat landscape caused by its diverse and fragmented nature. It explores novel on-node detection methods for Distributed Denial of Service (DDoS) attacks and dynamic network routing strategies to ensure packet safety through a trust-aware Moving Target Defence approach. To ensure the security of the 5G Airport enabled digital communication system, we also explore and develop novel on-node detection methods for DDoS attacks. Additionally, we employ dynamic network routing strategies to ensure packet safety through a viable deployment of trust-aware Moving Target Defence (MTD). MTD involves continuously changing system configurations and communication paths, making it difficult for potential attackers to identify and exploit vulnerabilities effectively. By combining cutting-edge 5G technologies with advanced security measures, the project aims to enhance the digital communication ecosystem in aviation, promoting safety, reliability, and resilience in the face of emerging threats.