For decades, video surveillance in facilities management has been a passive, “looking back” technology. We used it to investigate what had already happened—a slip-and-fall, a security breach, or a workflow bottleneck. But just as large language models (LLMs) like ChatGPT redefined our interaction with text, a new generation of video language models (VLMs) is fundamentally transforming how we use video.
Cameras have evolved from passive recording devices into intelligent observers capable of understanding visual context and executing complex instructions. For the facilities manager, this marks the transition from passive observation to active “video supervision,” unlocking a new frontier for safety and operational insight.
The ‘iPhone Moment’ for Facility Oversight
Tech leaders like Google and OpenAI have invested billions into training models that far surpass traditional video analytics. By bringing these models to the edge and the cloud, cameras can now achieve a level of situational reasoning that mimics common sense. We are seeing the industry experience its “iPhone moment”—a point where technology becomes so transformative that it enables entirely new categories of utility.
In this new era, three foundational assumptions are shifting:
- From Liability to ROI: Cameras are moving from reactive “insurance” tools into proactive supervisors that drive cleanliness, compliance, and workflow integrity.
- The Death of the Server Room: The need for expensive, specialized edge hardware is fading. Sophisticated cloud models provide high-level reasoning via standard cameras at a lower Total cost of ownership (TCO) than legacy on-premise systems.
- Scalable Context: While human monitoring fails to scale, AI supervision offers genuine contextual understanding—distinguishing between a delivery driver and a loiterer, or a temporary shadow and a genuine safety hazard.
The Supervision Engine: AI Orchestration
The path forward lies in the synergy between AI and cloud video. A well-architected platform aggregates video with multiple AI models to deploy curated “agents” without complex local infrastructure. Ideally, a vendor provides industry-specific templates (e.g., warehouse compliance or retail cleanliness) while allowing customization for unique use cases.
When AI systems follow rules independently, accountability is paramount. A dedicated AI orchestration layer allows FM teams to define governance by aligning AI triggers with corporate policies. This keeps control centralized and transparent, ensuring that the technology serves the facility’s specific safety and operational goals.
Solving Real-World Challenges
The value of AI-driven supervision comes from deploying it with intent. Visionary organizations prioritize use cases where video intelligence delivers unambiguous outcomes:
- Safety & Compliance: Real-time PPE adherence and fire-exit obstruction detection.
- Operational Excellence: Automated spill detection in high-traffic lobbies and monitoring workflow integrity on loading docks.
- Economic Efficiency: Cloud-based AI is becoming one of the most cost-effective options, as it is often applied only to specific cameras that require it. Factoring in power, maintenance, and IT overhead, these subscriptions frequently cost less than maintaining aging on-premise hardware.
The Roadmap: Implementing Supervision Without the Rip-and-Replace
Transitioning to AI-powered oversight is an evolution, not an overhaul. FMs can layer intelligence onto existing systems through a structured, three-step pilot:
- The Operational Audit: Identify friction zones—loading docks where safety protocols are skipped or high-traffic areas prone to spills. Evaluate if existing hardware can support cloud-direct AI or if bridge devices are needed.
- The Hybrid Pilot: Test alerting capabilities on a specific outcome, such as unauthorized access. This stage focuses on refining alert-fatigue thresholds to ensure notifications are always actionable.
- Co-Defining AI Agents: Collaborate with frontline staff to determine priorities. When staff see AI as a tool that reduces their manual reporting burden and increases their safety, adoption moves from a mandate to a shared success.
Building Trust in the Autonomous Era
As supervision expands, leadership must define clear ethical boundaries. Systems should focus on environments and processes rather than personal surveillance or opaque scoring of individuals. Transparency and alignment with corporate values make this scale possible.
The differentiator for FM leadership in the next decade will not be the speed of deployment, but the caliber of governance. By moving from retrospective review to real-time supervision, facilities can identify risks and optimize processes before they become liabilities. Trust will be built on standards for transparency, ensuring that as intelligence scales, human oversight is preserved.
Robert Messer is the CEO of IPTECHVIEW, a provider of cloud-based video surveillance and security solutions.