Mr. Mahendra krishnapatnam
Securing the Eyes of Robots: Why Vision Needs Cyber Protection
Abstract:
As robots become increasingly integrated into daily life—from autonomous vehicles to healthcare assistants—their ability to "see" through cameras and sensors is vital for safe and intelligent operation. However, these vision systems also introduce new cybersecurity vulnerabilities that are often overlooked. Attackers can manipulate visual data, hijack camera feeds, or inject adversarial inputs that cause robots to misinterpret their environment, putting safety, privacy, and trust at risk.
This session explores the critical need to secure robot perception systems from a cybersecurity and privacy standpoint. We will examine real-world risks, common attack vectors like deep fakes and spoofed visuals, and showcase simple, effective strategies such as encryption, access controls, real-time anonymization, and edge computing. The talk will also highlight a privacy-by-design approach to robotic vision, providing attendees with actionable insights to build safer, more responsible autonomous systems.