Asylon and NVIDIA Partner to Advance Robotic Physical AI
New AI-powered monitoring architecture integrates edge and cloud intelligence to enhance human-robot collaboration in real-world environments.
- By Jesse Jacobs
- Mar 24, 2026
Asylon has announced a collaboration with NVIDIA to develop next-generation artificial intelligence analytics for autonomous robotic security systems.
The partnership introduces DroneIQ Overwatch, a capability within the DroneIQ platform designed to advance "Physical AI." The system enables humans, robots and AI to operate simultaneously in live environments. Asylon, a member of the NVIDIA Inception program, expects to roll out the technology in 2026.
Centrally managed through a 24/7 robotic security operations center, the new software layer continuously analyzes live video streams and operational data. The system automatically surfaces anomalies and security events for human review.
Unlike traditional monitoring that relies on human observation, the architecture is designed to augment operators by serving as an initial layer of situational awareness. The goal is to identify potential threats faster while maintaining human accountability.
The system utilizes a hybrid edge-to-cloud architecture. Asylon robotic platforms incorporate NVIDIA Jetson modules for onboard processing, allowing for immediate inference at the "edge" where the robot is deployed. Simultaneously, cloud-based GPU infrastructure supports model orchestration and continuous refinement of the AI.
This framework allows for real-time detection, centralized model management and scalable analytics across entire robotic fleets. Asylon executives stated the transition toward Physical AI is a strategic move to address increasing security complexity and workforce constraints by operating intelligence natively within physical hardware.
Asylon plans to showcase the platforms and the future of the Overwatch system at the upcoming ISC West conference in Las Vegas.