A core pillar of Dr. Capraru's work explores how hackers can weaponize naturally occurring environmental phenomena to execute cyber-physical attacks. In his landmark 2024 paper, co-authored with researchers from Imperial College London and A*STAR, he proved that rainfall decreases the technical overhead required for bad actors to trick self-driving cars.

While once seen as "low-resolution" compared to LiDAR, modern radar—powered by —is proving to be the backbone of all-weather reliability. By using synthetic datasets and neural style transfers, we can now train algorithms to recognize objects through the "fog" of environmental interference. What's Next?

His transition to the Ph.D. program at NTU marked a significant evolution in his research scope. While his undergraduate work was sensor-centric, his doctoral research has grown to encompass the larger system-level security and reliability issues of those sensors. This shift is evident in his collaboration with leading experts like , a renowned figure in resilient information systems security. Supported by the NTU-Imperial Global Fellows Program , this collaboration has been instrumental in steering his research toward the critical area of cybersecurity for autonomous vehicles.

While Richard Capraru tends to keep specific client names confidential (a trait of high-end strategists), public records and industry case studies reveal patterns of his interventions. Let’s look at two archetypal scenarios he has navigated.

: Standard machine learning models are trained to classify what they see, meaning they struggle to differentiate between a naturally bounced photon and a precisely timed adversarial laser pulse. Weaponizing the Elements: LiDAR Spoofing in Adverse Weather

is an active cybersecurity and computer engineering researcher whose core work focuses on securing the perception layers of autonomous vehicles against sophisticated adversarial manipulation. His recent breakthroughs explore how bad actors can exploit physical environment anomalies, such as severe weather, to execute highly targeted LiDAR spoofing attacks . By examining the intersection of sensor hardware, machine learning perception models, and real-world environmental factors, Capraru's research provides critical blueprints for building the next generation of resilient autonomous systems. The Evolution of Adversarial Sensor Attacks