Secure Passage is integrating EyePop.ai’s customizable computer vision technology into Truman, its identity-first Physical AI operating system, to connect camera-based detection with identity data, access permissions, schedules, and operational workflows. Announced by the companies, the integration is designed to help enterprise security teams distinguish routine activity from potential threats and coordinate responses involving personnel, connected robots, or drones. The system can automate actions under customer-defined policies or require operator approval, depending on operational requirements.
Secure Passage Connects Computer Vision With Physical AI
Enterprise security systems generate substantial volumes of camera footage and sensor alerts, but identifying an object or movement does not necessarily explain whether an event is authorized or requires intervention. Secure Passage aims to address that gap by combining EyePop.ai’s visual intelligence with Truman’s identity and operational context.
The integration connects computer vision models with information such as employee and contractor credentials, access permissions, schedules, location data, and connected physical infrastructure. The objective is to move security operations beyond isolated alerts toward context-aware detection, investigation, and response.
For example, a person entering a restricted area might be an authorized employee, a scheduled contractor, or someone without permission. A camera can detect the entry, but determining whether it is anomalous requires additional information. Truman is designed to correlate the visual event with identity records and operational rules before initiating an appropriate response.
The companies describe this approach as Physical AI: connecting machine perception with contextual decision-making and coordinated action in real-world environments.
How the EyePop.ai Integration Works
EyePop.ai provides customizable computer vision models that organizations can adapt to specific sites and operational requirements. According to Secure Passage, potential applications include tailgating detection, perimeter monitoring, personal protective equipment compliance, and equipment-condition monitoring.
When EyePop.ai identifies a visual event, Truman can correlate the detection with connected access-control systems, identity providers, digital twins, and other operational platforms. That additional context can help security teams prioritize incidents instead of treating every camera alert as equally significant.
The combined platform is designed to support three main functions: identity-driven context, contextual anomaly detection, and AI-assisted investigation and response.
Identity-driven context links visual observations with available credentials and identity records. Anomaly detection compares observed activity against expected schedules, locations, and surrounding events. Investigation workflows can assemble supporting evidence and notify relevant personnel or initiate predefined procedures.
These capabilities depend on the accuracy and availability of the underlying identity records, integrations, camera feeds, and customer-defined policies. The companies have not disclosed independent performance benchmarks demonstrating how accurately the integrated platform distinguishes authorized activity from genuine security incidents.
From Visual Alerts to Coordinated Response
The integration’s broader proposition is its ability to connect detection with action. Under customer-defined policies, Truman can trigger notifications, escalate incidents, or coordinate connected robots and drones to verify events.
Organizations can configure responses to run automatically or require human approval. That flexibility is important in environments where an automated response could affect employees, visitors, physical equipment, or business operations.
For security teams, the potential benefit is reducing the manual work involved in monitoring cameras, correlating records, and conducting initial investigations. Rather than asking operators to review every alert independently, the system aims to supply contextual information and route incidents to the appropriate response process.
However, autonomous physical responses introduce requirements that go beyond conventional video analytics. Enterprises must consider access controls, audit trails, fail-safe behavior, privacy obligations, and clear escalation procedures. Human oversight remains particularly important when an event is ambiguous or a response could create safety risks.
Custom Models and Existing Infrastructure
EyePop.ai’s self-service model-creation approach is intended to help organizations configure computer vision for their own facilities and operating conditions. This could be relevant to manufacturing plants, logistics facilities, corporate campuses, and other environments where security requirements vary by location.
Secure Passage also positions Truman as a layer connecting existing cameras and operational systems, rather than requiring organizations to treat video surveillance as a standalone technology. The actual level of compatibility will depend on supported devices, available integrations, and deployment configurations; the announcement does not provide a complete compatibility list.
The companies said the integration is available immediately across Truman deployments. Secure Passage also offers support through its 24/7 Managed Operational Intelligence Services, operating from its Global Security Operations Center in Kansas City.
What It Means for Enterprise Physical AI
The partnership reflects a broader direction in enterprise AI: combining computer vision, identity systems, and workflow automation to make physical operations more responsive to context.
Traditional video analytics typically focus on detecting predefined visual events. Context-aware platforms seek to connect those observations with organizational data and operational rules, potentially reducing the gap between an alert and a decision.
For enterprise buyers, the important distinction is whether the technology can improve incident prioritization and response without generating excessive false alarms or introducing unacceptable operational risk. Buyers will also need to evaluate data retention, model accuracy, identity-data governance, and the reliability of automated actions.
Secure Passage and EyePop.ai are positioning their integration as a step toward that model. Its practical value will depend on how well the system performs in real-world deployments and whether organizations can demonstrate measurable improvements in response times, investigation workloads, and security outcomes.
Market Landscape
Enterprise physical security is increasingly intersecting with computer vision, identity management, edge computing, and AI-driven workflow automation. Video analytics can detect movement or predefined events, while identity-aware systems add information about who should be present, where they are authorized to go, and when access is expected.
Secure Passage and EyePop.ai are targeting this convergence through an integration that connects customizable vision models with Truman’s identity-first architecture. The approach sits within the broader Physical AI market, where AI systems interact with real-world environments through connected devices, sensors, robots, and operational platforms.
The competitive landscape spans video surveillance analytics, physical access control, security orchestration, and robotics platforms. Established security vendors and emerging AI providers address different parts of this ecosystem, making integration quality and operational reliability important considerations for buyers.
The announcement does not include independently verified accuracy figures, customer deployment metrics, or comparative performance results. Enterprise adoption will therefore depend on validation under site-specific conditions, alongside governance controls for identity data and autonomous responses.
Top Insights
- Secure Passage is integrating EyePop.ai’s customizable computer vision into Truman to connect visual detections with identity records, access permissions and operational context.
- The combined platform is designed to support tailgating detection, perimeter monitoring, PPE compliance and equipment-condition monitoring across enterprise environments.
- Truman can initiate policy-driven workflows, including notifications, incident escalation and coordination with connected robots or drones.
- Organizations can configure responses to run automatically or require operator approval, allowing workflows to reflect different operational risks and requirements.
- The integration is available across Truman deployments, but independent accuracy benchmarks and quantified customer outcomes have not been disclosed.
Power Tomorrow’s Intelligence — Build It with TechEdgeAI
