Dr Adolfo Perrusquía - Cranfield University

Increased proliferation of drones and autonomous systems can disrupt critical national services. Whilst many are amateur drones that pose no malicious intention, some may carry deadly capability and cause severe economic damage to the UK’s infrastructure. Protection against malicious drones is critical to ensuring smooth operation of services, whilst safeguarding it against the most severe threats. As drone intentions are often hidden and false positives need to be low, radical innovation in machine learning and inference is required.
My current research aims to design innovative and stable machine learning algorithms by exploiting experts’ aerospace knowledge and deep patterns in a complementary mechanism that can not only achieve high intention prediction accuracy, but also low and stable false positive rates – essential for protection of key economic targets. This will benefit the UK economy, protecting it against potentially malicious drones, but also not overreacting to diverse low-flying objects.