Emesent, founded in 2018 with roots in Australia's CSIRO, builds autonomous drones and spatial computing systems that operate where humans can't - or shouldn't. The threat model here is physical: GPS-denied mines, collapsed structures, hazardous industrial sites where a machine's ability to navigate, avoid collisions, and generate high-fidelity 3D maps replaces putting people at risk. Their flagship product, Hovermap, is a mobile LiDAR scanning unit that combines autonomous flight with advanced collision avoidance, deployed across 40 countries in underground mines and construction sites.
From a security perspective, the attack surface is non-trivial. These are AI-driven robotic systems making real-time navigation decisions in environments with zero margin for error - autonomous flight control, sensor fusion, spatial mapping pipelines. The company converges artificial intelligence, autonomous robotics, and advanced spatial computing, which means the engineering stack spans embedded systems, computer vision, SLAM algorithms, and the data infrastructure that processes point clouds from hazardous environments. A 100-plus person team operating at that intersection of physical autonomy and perception software faces the usual suspects: firmware integrity, sensor spoofing, data pipeline security, and the integrity of autonomous decision-making under adversarial conditions.
For anyone interested in the security implications of physical AI systems - where a compromised sensor or corrupted map doesn't just mean bad data but potentially a drone collision in a mine shaft - Emesent represents a concrete, high-stakes domain. The company's work sits at the convergence of cybersecurity and operational technology in its most literal form: machines that must be trustworthy because they operate autonomously in environments designed to be hostile to both humans and electronics.
