Dev.to Security πŸ” Cybersecurity πŸ‘ 0 πŸ“– 10 min read

Before You Replace the Camera, Test the Recording

A camera can be connected, recording, and correctly identified in a video management system while still failing to provide the information its operator needs. A clear live image does not necessarily mean the recorded foo

A camera can be connected, recording, and correctly identified in a video management system while still failing to provide the information its operator needs. A clear live image does not necessarily mean the recorded footage will support an investigation, and a healthy device status does not establish that its original coverage requirements are still being met.

Liverpool John Lennon Airport's CCTV modernization offers a useful opportunity to examine that distinction. On November 11, 2026, Telent announced a partnership with the airport to assess its surveillance infrastructure, covering camera inventory, performance, and coverage across key operating areas. The airport operates approximately 500 cameras, and the assessment is intended to inform which parts of the existing system should remain, receive corrective work, or be replaced.

The work follows a three-month trial of Telent's Arbitex platform, during which daily checks examined connectivity, camera alignment against reference images, and image-quality conditions such as dirt or obstruction. Telent reported in May that the contracted rollout had started on April 1, 2026, following the trial. The November announcement broadens the assessment but does not establish the rollout's current completion status, the number of cameras included in each stage, or quantified improvements in downtime and maintenance costs.

For engineers responsible for large CCTV estates, the real challenge is turning those monitoring results into reliable decisions about the useful life of each camera.

A camera's job should be defined by what the recording can prove

Consider a service entrance monitored by an existing IP camera. During the day, the operator can see delivery vehicles approaching and leaving, and the VMS shows a continuous recording. At night, however, the security team may need to establish whether a person crossed a marked boundary before or after a vehicle entered the area.

As a truck approaches, its headlights create glare across the relevant part of the image. The recording shows movement near the entrance, but the order of events cannot be established confidently. The camera is functioning, and its field of view may not have changed at all, yet the recording falls short of what the operator needs for that particular investigation.

This is a hypothetical commissioning scenario, not a reported problem at Liverpool John Lennon Airport. It illustrates why image quality must be evaluated against an operational task rather than device specifications alone.

If the camera was installed primarily to give staff a general view of arriving vehicles, the existing footage might still satisfy the approved requirement. If the owner now expects the recording to establish the sequence of movements around a restricted entrance, the acceptance criteria need to reflect that additional responsibility.

IEC 62676-4:2025 provides application guidance covering video surveillance design, testing, commissioning, and maintenance. Axis Communications also explains why pixel density should be treated as a simplified model of image detail rather than a complete measure of surveillance performance. Lighting direction, optical quality, compression, and other conditions affect what a recording can actually reveal.

A subject occupying too few pixels may justify reviewing camera position, lens selection, or resolution. Glare and motion blur require a different investigation. Establishing the order of events also depends on recorded frame rate and dependable timestamps. These are different variables, and replacing the camera without identifying the limiting factor may not solve the problem.

The starting point should therefore be a requirement that can be tested from the recording itself: can an authorized operator establish the required event or action under the conditions the camera is expected to handle?

The reference image is useful only when its purpose is documented

A reference image gives maintenance personnel a known view to compare the current camera image against. If the camera moves after maintenance, is obstructed, or gradually loses its intended framing, that comparison can reveal a change that might otherwise go unnoticed.

The three-month Arbitex trial at Liverpool included alignment checks against reference images, and Telent says its personnel worked with the airport's technical team and on-site maintenance contractor to calibrate those references and refine the reporting parameters. This is a practical starting point for maintaining a large estate, but a matching image is only one part of proving that the view remains suitable.

A useful baseline should connect the approved camera view to the task it supports. That means documenting the relevant area, the activity operators need to observe, and the conditions under which the requirement must be satisfied. Where lighting or movement affects that task, representative recorded clips provide additional evidence that a single reference frame cannot capture.

For a service entrance, the acceptance material might include daytime and nighttime recordings of representative vehicle movements, together with the recording profile and the date of the test. For a view used in investigations, the accepted footage should be played back through the tools available to the duty team, not evaluated exclusively from the camera's configuration interface.

This record also makes future maintenance decisions more consistent. A slight shift at the edge of a general overview may have little operational consequence, while a smaller movement near a critical doorway could remove the only usable view of an important action. The maintenance technician needs to understand that difference without reconstructing the original design intent.

Reference images also require change control. Renovations, changes in operational use, or new obstructions may justify updating a camera's approved view, but the replacement baseline should reflect a confirmed requirement rather than merely accepting the camera's current position. Otherwise, refreshing the reference can make the monitoring result appear normal without resolving a loss of useful coverage.

Test the recording path, not just the camera output

A camera's live image is only one stage of the video system. The recording that reaches an investigator may pass through different stream settings, compression, storage, and playback processes before the operator can review it.

Some VMS configurations allow different streams for live viewing and recording. Milestone XProtect's documented multi-streaming functionality provides one example of this distinction, although the published Liverpool Airport accounts do not identify the site's VMS or confirm that its streams are configured this way.

The practical consequence is straightforward. An operator may see a satisfactory image in the live grid while the stored footage contains less detail than the assigned task requires. Before concluding that a camera needs replacement, the maintenance team should establish whether the problem exists at capture, in the configured recording stream, or elsewhere in the recording and playback chain.

A controlled test can compare the current field of view with an accepted recording captured under relevant conditions. The technician can then review the recorded stream's configuration, playback quality, timing, and any changes made to the camera or VMS since acceptance. The objective is to identify where useful information is being lost, rather than assuming that the device at the wall is responsible.

Export should also form part of the test wherever sharing evidence is an expected operational task. An authorized operator should be able to retrieve the relevant recording through the approved process and confirm that the exported material can be opened and examined with the detail required for its purpose.

This matters particularly for investigations involving several cameras. Even when each recording is individually usable, unreliable timestamps can make it difficult to reconstruct the relationship between events. A surveillance system intended to establish the order of activity needs to be evaluated as a recording and retrieval system, not merely as a collection of image sensors.

A camera replacement can improve capture quality without correcting an unsuitable recording profile or an unresolved playback problem. Testing the complete path first gives the owner a better chance of funding the correction that actually changes the outcome.

Retain, correct, or replace: three different engineering decisions

Once the required performance is documented and the recorded evidence is available, the modernization decision becomes more defensible. Not every camera installed for several years needs replacement, and not every technically functioning camera should remain in service.

Retain is appropriate when the camera still performs its approved task, provides acceptable recorded results, and has a supportable operating lifecycle. That assessment includes more than optical performance. Software maintenance, security updates, and compatibility with the planned VMS environment also affect whether an otherwise useful device remains a sound choice.

Correct applies when a deficiency can reasonably be addressed without replacing the device. Restoring camera aim or focus, removing an obstruction, adjusting lighting, or correcting recording configuration may recover the required performance. Verify the result using the same operational task that identified the deficiency, and retain the original and corrected recordings as part of the maintenance evidence.

Replace becomes justified when the device cannot satisfy the approved requirement through practical correction, or when support and compatibility limitations make continued operation unsuitable. A different lens, mounting position, or additional dedicated camera may also be needed when the existing view cannot reliably perform two incompatible jobs.

These decisions should not be driven by the model number alone. Two cameras of the same age may have very different operational value depending on their location, recording requirements, maintenance condition, and available support.

For the hypothetical service entrance, that means investigating the cause of the nighttime problem before selecting new hardware. A framing adjustment may restore the relevant boundary, or the limiting factor may lie in lighting, exposure, or recording settings. If the required sequence of events still cannot be established after those conditions are addressed, the evidence supports a more substantial design change.

The outcome should be a documented decision linked to the particular view and its purpose. This allows an owner to retain equipment that is still suitable while directing modernization funds toward verified gaps rather than broad replacement assumptions.

Monitoring alerts should lead to verified restoration

Continuous or routine camera checks can help maintenance teams identify changes across a large CCTV estate. The value of those checks, however, depends on what happens after a condition is flagged.

A report that a view differs from its reference image is not necessarily proof of a defect requiring corrective work. Rain, headlights, scheduled lighting changes, and temporary activity may alter the image while leaving the camera suitable for its assigned task. Conversely, some deteriorating views may not generate a flag under the selected monitoring parameters.

An effective assurance process should therefore measure the quality of its decisions as well as the number of conditions it detects.

One useful approach is to review a sample of flagged events to determine how many actually required corrective action. A second sample, drawn from views that produced no flags under comparable conditions, can help identify problems that the checks missed. Together, those reviews provide a basis for refining monitoring parameters without assuming that fewer alerts always mean better coverage.

The maintenance record should also follow a confirmed issue through to its resolution. The initial condition, verification, corrective action, and final recorded test belong to the same operational history. Closing a ticket because a camera has reconnected or its reference image now matches is not the same as demonstrating that the camera has recovered its required function.

For a larger deployment, starting with a manageable group of representative views can make this process easier to evaluate. The team can compare how different defects are identified, which require a site visit, which can be corrected through configuration or routine maintenance, and how consistently the final recording meets the approved task.

Where comparable historical records exist, time to verified restoration may be more informative than raw alert volume. It measures the interval between identifying a confirmed deficiency and demonstrating that the required camera function has been restored.

Telent's published Liverpool accounts do not provide measured improvements in downtime, restoration time, or maintenance cost. Those remain reasonable performance measures to consider for a similar project, but they are not results that can be attributed to the airport's deployment.

The most valuable upgrade deliverable is a defensible decision

A large CCTV estate can accumulate different camera generations, mounting arrangements, recording profiles, coverage requirements, and maintenance histories. Replacing every older device may be expensive without resolving the most important surveillance problems, while retaining cameras solely because they remain online can preserve weaknesses that operators have learned to work around.

Liverpool John Lennon Airport's modernization illustrates an approach that begins with understanding the existing system. Inventory, condition monitoring, reference images, and coverage assessment can provide useful evidence before decisions about replacement are made. The published material does not yet establish the final outcome of that process, but the engineering opportunity is clear.

For each camera, the owner should be able to understand its assigned function, the recording used to verify that function, any observed deficiency, and the reason for retaining, correcting, or replacing the device. That information should survive beyond the original project team so future maintenance staff can repeat the relevant checks after repairs, configuration changes, or building modifications.

A camera can be physically healthy and still produce unsuitable evidence. It can also be old and continue performing its approved task reliably. The distinction becomes visible only when the system is tested against what people actually need to see and establish from the recording.

The strongest modernization decision is not based on how new the camera is. It is based on whether the recorded result still does the job.

Sources and further reading

AI disclosure: This article was drafted with AI from a human-reviewed, primary-source-verified technical master.

πŸ“° Read the original article on Dev.to Security

Originally published by Dev.to Security. Aggregated on AIWithGhost for educational purposes β€” full credit and traffic to the original publisher.