The first article in this series argued that operational visibility has become fundamental to how modern data centers operate. The second showed why that visibility depends on security that is layered, integrated, intelligent and cybersecure.

This third article moves from architecture to outcome. When video, access control, network audio, analytics and environmental monitoring work as one connected data center security architecture, they can provide more than footage. They can deliver timely context that helps teams detect, verify and respond before a small event becomes operational disruption. That is the shift from surveillance as a system of record to surveillance as operational intelligence.

Surveillance is no longer just a system of record

For much of its history, video surveillance in data centers was mainly retrospective. Cameras recorded activity, and teams reviewed footage after an event. That evidentiary role remains essential, but it is no longer enough for facilities that operate continuously across more zones, access points and connected systems.

Edge-based analytics and connected sensors can now identify defined events in real time – such as unexpected movement in a restricted area, tailgating at a controlled doorway or environmental conditions that require attention – and trigger the relevant workflow. The goal is not to replace human judgment. It is to help operators see the events that matter, verify them faster and respond with better context.

Safety is also a visibility problem

Modern data centers contain high-voltage electrical rooms, battery systems, cooling infrastructure and mechanical plants, while contractors, vendors and maintenance teams may be working across several zones at once. Engineers may also work alone in data halls, where a fall or medical event could otherwise go unnoticed. In that environment, safety depends partly on knowing who is present, whether access is authorized and what is happening around critical equipment.

Integrated video and access control can help confirm access events, support remote assessment, alert teams to unexpected presence, and create a reliable record for investigations and audits. With suitable Axis and partner analytics, audio can add another layer of awareness by detecting distress sounds, such as screams or calls for help, or other unexpected sounds in areas where camera views are limited. Network audio can then support targeted warnings or emergency instructions.

Together, these capabilities make physical security part of the operating response, rather than a separate system reviewed only after an incident.

"Data center leaders should no longer evaluate surveillance and access systems solely as security investments. Across APAC, operators increasingly expect them to enable earlier detection, faster and more consistent incident response, and stronger safety and operational visibility across an entire portfolio. At multi-country scale, this requires a common operational view, consistent service support, sound IT governance and systems that local teams can manage effectively; simply adding devices to fragmented infrastructure increases complexity without necessarily improving outcomes," Wen Bin, sales director, SEA & South Korea at Convergint.

"The strategic priority is therefore to define the required business, resilience and regulatory outcomes first, then build a phased roadmap that aligns architecture, budget, skills and support around achieving them."

Operating a portfolio, not only a building

Many hyperscale and colocation operators manage facilities across multiple cities or countries. Their central teams need consistent alarm handling, access workflows and evidence across the portfolio, while local teams still need the flexibility to meet site-specific operating and regulatory requirements.

This makes interoperability a strategic requirement. Teams should be able to view live and recorded video, correlate access events, investigate incidents remotely, and route verified alerts to the right people without rebuilding the process for every site. Standardization should reduce operational variation without forcing every facility into an identical design.

The efficiency dividend

Once video, access and sensor data are connected to operational workflows, the same infrastructure can support outcomes beyond security. It can reduce manual verification, help teams prioritize alarms, avoid unnecessary dispatches and shorten investigations. Automation extends that value: biometric access workflows can streamline verification, while video analytics and connected sensors can provide inputs on conditions such as temperature, condensation, water levels or leaks to building management and IoT systems for faster action. The same data can also reveal recurring exceptions – for example, repeated access issues or areas generating excessive nuisance alarms – that deserve a process or design change.

That creates a clearer business case. Rather than counting cameras or alerts, operators can measure time to verify, false-alarm rates, response completion, investigation time and the quality of audit evidence. These are operational measures that connect technology investment to resilience and efficiency.

Why an open, integrated architecture matters

Data centers rarely operate as single-vendor environments. Axis and its technology partners can bring together network video, access control, audio, radar, thermal imaging, analytics and environmental monitoring through an open, IP-based architecture. Edge analytics prioritize and process relevant events at the source to reduce bandwidth use and latency, while hybrid integrations connect those events to cloud or on-premises platforms and the workflows operators already use.

The architecture must also be secure throughout its lifecycle. Connected physical security devices need strong authentication, encrypted communications, controlled privileges, network segmentation, signed firmware, secure boot and a disciplined approach to updates and maintenance. A lifecycle approach to cybersecurity is what allows new functionality to be introduced without weakening the environment it is intended to protect.

Integration is what turns data into intelligence

Integration does not mean sending every data point to every system. It means designing the right connections around specific decisions. A door event might be paired with video for immediate verification. An environmental alert might open the nearest camera view and notify facilities staff. An incident record might combine the relevant access, video and operator actions into one audit trail.

When those workflows are defined in advance, teams can move from reconstructing incidents after the fact to making informed decisions while an event is unfolding. That is the practical mechanism behind operational intelligence.

"In modern data centers, the evolution of edge AI is turning physical security infrastructure from a passive system of record into an active source of operational intelligence. Advances in on-camera AI processing and flexible edge application platforms now allow sophisticated AI solutions, such as CyberLink FaceMe facial recognition, to run closer to where security events occur, with less reliance on centralized GPU infrastructure," says Mei Guu, president at CyberLink.

"This enables practical applications such as facial recognition-based access control, two-factor authentication, visitor identity verification, restricted-area protection and personnel accountability during emergencies. The next challenge is connecting this intelligence across video, access control, video management systems (VMS), environmental and other operational systems. While standards such as ONVIF Profile M have advanced interoperability, metadata and semantic consistency across multi-vendor environments are still evolving."

From intelligence to repeatable execution

The next step is not to add technology for its own sake. It is to begin with the operating decision: what must be detected, who must verify it, what context they need and what action should follow. It also requires deciding where that intelligence should be used: in the VMS for security workflows, or in building management, IoT and access control platforms when Axis devices are used as operational sensors. From there, operators can build a connected architecture that is measurable, secure and repeatable across sites.

The final article in this series will examine how that shared strategy must be adapted across APAC markets with different regulations, maturity levels and customer expectations.

For a practical view of the technologies and five layers of protection involved, download the free Data Center Security eBook.