Verobotics deployed AI-powered façade robots at NVIDIA's Israel campus, capturing 20,000 images and identifying 40 anomalies across 100,000 sq. ft. of building envelope

Verobotics combined robotic façade cleaning, AI vision, and edge computing on NVIDIA Jetson hardware at the NVIDIA Israel campus, covering roughly 100,000 sq. ft. of building envelope and 3,000 windows and façade sections. About 60% of façade cleaning was completed robotically; about 40% was completed with traditional cleaning support.

The robot processed approximately 20,000 façade images onboard using edge AI (processing data locally on the robotic platform rather than relying on cloud connectivity), enabling faster interpretation of changing conditions during live operation in unpredictable façade environments with glare, shadows, reflections, and wind.

During the deployment, Verobotics identified 40 façade anomalies requiring escalation and engineering review. The system created a large-scale visual dataset enabling longitudinal façade analysis, anomaly detection, deterioration tracking, and eventually predictive maintenance models for high-rise buildings, shifting the workflow from reactive maintenance to continuous architectural intelligence.

The deployment reduced human work-at-height exposure while demonstrating what commercially viable robotics looks like: a hybrid operating model combining robotics, human crews, and AI-assisted inspection workflows integrated into routine building operations rather than forcing full automation.