Ghost plates are no longer a peripheral concern. Their use has been identified in connection with serious and organised crime, and their effectiveness against standard ANPR infrastructure makes them a rational tool for those who need to operate outside the national intelligence radar.

This paper examines the operational risks ghost plates create, the limitations of traditional ANPR systems in detecting deliberate manipulation, how AI-enabled analytics can identify plate anomalies in real time, and the findings from a live operational trial conducted with Surrey Police.

1.ANPR as Critical Policing Infrastructure

ANPR is embedded across UK policing operations. It supports intelligence-led vehicle stops, cross-force data sharing, suspect vehicle tracing, pattern-of-life analysis, and post-incident investigation. It is among the most widely used and data-rich tools available to operational policing.

The system rests on a foundational assumption: that vehicles moving through monitored corridors generate reliable, actionable data. When a vehicle passes through undetected, not due to equipment failure but as a result of deliberate plate manipulation, that assumption no longer holds. The gap it creates is invisible by design, and consequently difficult to quantify or address.

Even a small proportion of manipulated plates can introduce disproportionate risk to intelligence workflows. A vehicle of interest that consistently evades ANPR across multiple force areas does not simply avoid detection at a single point. It is removed from the national intelligence picture altogether, with no record of its absence to prompt further enquiry.

2.The threat: Ghost Plates as deliberate evasion

Ghost plates are designed to exploit the optical and infrared limitations of camera systems. They may reflect or absorb infrared light in ways that prevent character capture, distort legibility under specific lighting conditions, or appear fully compliant in daylight while failing under ANPR illumination.

Those manufacturing and using these plates have a working understanding of how ANPR systems operate. They are aware of the technical thresholds the systems are built around and are constructing plates specifically to defeat them. This represents targeted exploitation of a known vulnerability, not incidental or accidental non-compliance.

For policing, the consequence extends well beyond a failed read. These plates are associated with the deliberate creation of blind spots within national intelligence infrastructure. Vehicles using them have been linked to drug supply networks, vehicle-enabled acquisitive crime, disqualified and uninsured driving, and evasion of targeted vehicle markers.

The Surrey Police trial that underpins this paper provides operational evidence of the scale of the problem. Within the trial period, multiple vehicles displaying ghost plate characteristics were identified in a single corridor. Each of those vehicles, prior to the trial, would have passed through that corridor without generating any intelligence record.

3.Limitations of traditional ANPR in detecting manipulation

Conventional ANPR systems are optimised to read compliant plates under expected operating conditions. They are not configured to identify abnormal reflectivity patterns, flag unreadable plates as potentially suspicious events, analyse plate geometry against DVLA standards, or correlate visual anomalies with registered vehicle data.

When a ghost plate passes through an ANPR camera, the system typically logs a failed read and proceeds. No alert is generated, no intelligence record is created, and no further action is prompted. The vehicle exits the monitored zone without trace.

The core vulnerability is that the system treats deliberate evasion identically to a dirty plate or a momentary camera obstruction: as a data quality issue rather than a threat signal. As awareness of this gap spreads among those who benefit from exploiting it, its use is likely to grow, while remaining invisible to the systems responsible for detecting it.

In intelligence-led policing, this kind of silent gap is particularly difficult to manage. Unlike a visible system failure, it produces no error logs, no alerts, and no prompts for review. There is nothing to indicate that anything has been missed.

4.How NOW Wireless identifies Ghost Plates

NOW Wireless has developed an AI-enabled analytics capability designed to operate alongside existing camera and ANPR infrastructure. Rather than replacing ANPR, it augments the existing detection model to identify vehicles that are attempting to defeat it.

The system analyses video feeds in real time, applying trained models to identify characteristics associated with ghost plate behaviour: abnormal infrared reflectivity signatures, plate geometries inconsistent with DVLA-standard formats, and discrepancies between the visible plate format and the registered vehicle class or make.

Where a conventional ANPR system logs a failed read and takes no further action, the NOW Wireless system identifies that read as a potential anomaly, cross-references available vehicle data, and, where confidence thresholds are met, generates an operational alert.

The system is designed to support officer decision-making rather than replace it. Alerts are accompanied by imagery and the analytical parameters that generated them, giving operators the information required for a rapid, informed assessment. Enforcement decisions remain with the officer. The system’s role is to surface intelligence, not to determine outcomes.

This distinction matters both operationally and from a governance perspective. AI outputs that are explainable, auditable, and clearly positioned as decision-support tools are more appropriate to policing environments than systems that assert conclusions without providing the supporting basis.

5.Operational trial: Surrey Police

In a live operational trial with Surrey Police, the NOW Wireless AI analytics capability was integrated with existing CCTV and ANPR infrastructure. The trial was designed to assess real-world detection performance under operational conditions, without modification to the underlying camera network.

Metric	Outcome
Detection speed	Alert generated within approximately 30 seconds of plate anomaly identification
Enforcement outcomes	16 during the trial period
False positive rate	Zero recorded during the trial
Infrastructure disruption	None — integrated with existing CCTV and ANPR hardware
Officer deployment	Immediate — real-time alerts enabled proactive intervention

The trial demonstrated that ghost plate detection can be operationalised within existing infrastructure, at speeds compatible with real-time intervention, and without generating false positive rates that would undermine officer confidence or create unsustainable operational demands. The number of enforcement outcomes is significant, but secondary to what the trial established: that deliberate ANPR evasion can be systematically identified and acted upon within a live operational environment.

Each of those vehicles, had the trial system not been in place, would have exited the monitored zone without generating an alert, an intelligence record, or a deployment. The value of the trial lies not only in the outcomes it produced, but in the operational picture it revealed.

6.Operational value to UK Police Forces

Strengthening proactive policing capability
Real-time anomaly alerts enable officers to intervene before vehicles exit monitored zones. In high-risk corridors, on major arterial routes, and across known crime transit routes, this capability converts silent data gaps into actionable intelligence, at the point where intervention remains possible.

Protecting intelligence integrity
Treating unreadable or anomalous plates as risk events rather than data quality failures ensures that deliberate evasion enters the intelligence record. Over time, this enables pattern analysis: identifying vehicles that repeatedly trigger anomaly alerts across multiple locations, even where individual incidents do not reach the threshold for enforcement action.

Supporting evidential standards
The system preserves original imagery, logs the analytical parameters behind each alert, and maintains audit trails designed to support disclosure and evidential scrutiny. In prosecution contexts, the ability to demonstrate how an alert was generated, and on what basis, is as important as the alert itself.

7.Governance, integration and deployment

Forces evaluating enhanced detection capability will need to assess integration pathways with existing ANPR and control room systems, data security architecture and GDPR compliance, the explainability of AI outputs for operational and legal purposes, and scalability from pilot to force-wide deployment.

The NOW Wireless system is designed to layer onto existing infrastructure rather than require full network replacement. Deployment can be configured as edge-based analytics at camera level, centralised secure processing, or a hybrid architecture, depending on force requirements and existing estate.

The Surrey trial demonstrated that integration is achievable without capital disruption to existing ANPR infrastructure. The more substantive considerations for forces concern governance and oversight: ensuring that AI-generated alerts are handled within clear operational frameworks, and that the system’s outputs can withstand legal scrutiny.

NOW Wireless is able to support forces through the governance assessment process, including providing technical documentation, explainability reports, and trial data to inform internal legal and information assurance review.

8.Strategic considerations for Police Leadership

As policing becomes increasingly data-dependent, adversarial adaptation is an expected consequence. Where criminal actors identify technical vulnerabilities in enforcement infrastructure, exploitation typically follows, often at a pace that outpaces institutional response cycles.

Ghost plates represent a current example of that pattern. The technology to defeat ANPR is available, its effectiveness is understood by those who use it, and its use is growing. The relevant question for police leadership is whether current systems are capable of reliably identifying deliberate plate manipulation, and, if not, what the intelligence cost of that gap represents.

A structured pilot in a high-risk corridor provides an evidence-based route to answering that question. Key metrics should include detection accuracy, false positive rates, time to operational response, the intelligence value of flagged vehicles assessed through post-stop outcomes and PNC checks, and measurable impact on proactive policing outcomes.

The Surrey trial provides a template for how that evaluation can be structured. Forces interested in conducting a comparable assessment can engage NOW Wireless to scope a pilot calibrated to their operational environment and infrastructure.

9.What a Pilot Evaluation involves

For forces considering a structured evaluation, the practical requirements are modest. A pilot typically requires identification of a suitable operational corridor with existing ANPR and CCTV coverage, a defined trial period of sufficient duration to generate statistically meaningful detection data, a nominated operational lead and data protection officer for governance oversight and agreed metrics against which the trial will be assessed.

NOW Wireless manages system integration, technical configuration, and ongoing analytical support throughout the trial period. Forces are not required to procure new hardware or modify existing infrastructure. Data generated during the trial remains subject to existing force data governance frameworks.

Pilot timelines typically run six to twelve weeks from integration to initial results, depending on the complexity of the existing infrastructure environment. Full evidential and governance documentation is provided as standard.

To discuss a pilot evaluation, contact NOW Wireless at nowwireless.co.uk