NEWS 05 November 2026

The speed-camera figures hiding in plain sight: 2,442 detections on one road, zero on another—and a camera dormant since 2019

Published 5 November 2026 at 23:29

NI NEWS

A comparison of PSNI freedom-of-information disclosures reveals what camera statistics can prove, what they cannot, and why headline speeding totals risk confusing enforcement activity with driver behaviour.

Five locations in the Derry area appear in a Police Service of Northern Ireland freedom-of-information response about speeding. On one, the A6 Glenshane Road, the police recorded 2,442 detections for the period asked about. At another, Crescent Link, the figure was zero. The difference is striking. It does not, however, mean that one route was demonstrably dangerous and the other free of speeding. The disclosure gives the outcomes of enforcement records; it does not say how many vehicles passed each location, how often cameras were operating or how many hours were observed.

That distinction becomes more important when the records are placed beside other published answers. A separate FOI records 106 mobile camera-van deployments and 463 speeding offences on Ravernet Road over three years. Another reveals that a fixed camera on the A24 Saintfield Road had last been operational in July 2019, although it was still physically present when the request was made in January 2025. A 2026 FOI contains only seven PSNI detections of drivers travelling at least 30mph in a 20mph zone over an 11-month period in 2025—but explicitly excludes the Northern Ireland Road Safety Partnership.

Read independently, each figure lends itself to a simple headline. Read together, they tell a different and more useful story: speeding data is fragmented across enforcement organisations, locations, recording categories and periods. A resident looking for evidence that a particular road is safer after a camera is installed cannot obtain that answer by counting tickets alone.

The five-site Derry disclosure

The first document, released under PSNI reference FOI-13942 on 20 March 2025, concerned five named sites and the value of fines paid during the requested 2023–24 financial-year period. The answer identified the locations more precisely than the questions did. The site described by the requester as Rossdowney Road was recorded as Dungiven Road; the location near a bank at Altnagelvin was the A6 Glenshane Road. Those corrections matter because a different road label can produce an entirely different result when somebody later searches public records.

The released table showed 967 detections and 346 fines paid on Dungiven Road, with a total paid value of £20,760. Duncastle Road at Newbuildings recorded 22 detections, 11 fines paid and £660. A6 Glenshane Road recorded 2,442 detections, 980 fines paid and £58,800. Limavady Road had 14 detections, four fines paid and £240. Crescent Link was listed as zero for detections, fines paid and paid value.

The five sites therefore do not simply form a league table of speeding behaviour. Their detection totals can reflect whether, when and how equipment operated. Without deployment hours and traffic flows, there is no sound way to compute a per-vehicle offence rate. The apparent contrast between 2,442 and zero is powerful precisely because it raises a question that the released table does not answer: how much opportunity was there to detect an offence at each site?

Why detections and paid fines are not the same thing

The A6 Glenshane Road's 980 fines paid represent approximately 40% of its 2,442 detections, a calculation based directly on the published figures. It would be wrong to call the remaining 60% 'unpaid fines' or 'drivers who escaped punishment'. A detection may follow different administrative routes, and the published table does not reconcile every recorded detection to a final outcome. Some cases may be eligible for an education course; others may be referred onward or processed on a different timetable. The document does not provide that detailed disposition.

There is a similar trap in the £58,800 received figure. It is a published monetary total associated with the fines-paid column. It is not an audited total of all revenue attributable to that location, and it is not proof that the camera exists to raise money. Financial reporting would require fees, costs, distributions and the final disposition of every case. The responsible public-interest question is whether enforcement is demonstrably reducing risky speeds and injury collisions—not whether an arithmetic multiplication can make a dramatic gross revenue headline.

Another limitation is that these are site records. A driver caught by a camera at one location may travel through several other roads without being monitored. Conversely, a road with a high detected total may be an intensively monitored route rather than one with a uniquely high proportion of speeding drivers. Enforcement numbers are useful for measuring the recorded workload and outcomes; they are much weaker as direct measures of community-wide driving behaviour.

What the Ravernet Road records add

A second disclosure, FOI/16341, published in December 2025, provided an unusually helpful additional denominator. Between 1 October 2022 and 30 September 2025, police said the Northern Ireland Road Safety Partnership had made 106 mobile camera-van deployments to Ravernet Road, detecting 463 speeding offences. The response grouped those detections by measured speed: 241 at 30–39mph, 211 at 40–49mph and 11 at 50–59mph.

Dividing the total detections by the number of deployments gives approximately 4.4 detections per deployment. That is a legitimate description of the disclosed enforcement yield. It is not a detection rate per hour or per vehicle, because the response does not give the duration of each deployment, the number of vehicles observed or the traffic speeds of motorists who did not trigger an offence. If some deployments lasted longer than others, a simple per-deployment average could conceal substantial differences in exposure.

The distribution also says very little about speed compliance without knowing the exact applicable limits, site setup and enforcement criteria for the recorded events. Although the speeds are grouped into bands, it would be unsafe to assume every one came from a single uniform 30mph limit unless the document explicitly establishes that. Eleven offences in the 50–59mph band are serious-looking numbers, but the dataset itself does not quantify the resulting collision risk or show how traffic speed changed once camera activity ended.

A fixed camera that was still there, but no longer working

The third document is more direct. Under FOI/00013618, a member of the public asked about a fixed safety camera on the A24 Saintfield Road, near Lenaghan Drive. The PSNI said that, as of the January 2025 point examined by the request, the camera was not operational and its most recent period of operation had ended in July 2019. When asked why it was still erected, the police said they were unable to explain why it remained in place.

There is an essential qualification: the disclosure distinguished that specific fixed camera from average-speed cameras on the same road. The presence of an inactive device cannot be treated as proof the whole corridor lacks speed enforcement. Nor does the response establish whether the physical camera has subsequently been removed or reactivated. What it establishes is that, at the date of the disclosure, the visible presence of one particular camera was not evidence that it was detecting offences.

This matters to drivers and residents for opposite reasons. A resident may assume a visible camera is offering continuous monitoring when it is not. A driver may assume all enforcement is absent when other equipment or mobile deployments exist. The case shows why public safety information needs equipment status, site type and dates—rather than a map of camera-shaped objects without an operational history.

Seven detections in 20mph limits: the comparison that should not be made

The most counterintuitive figures come from FOI/16823, published on 3 July 2026. The requester asked about the highest speeds and numbers of serious speed-limit breaches on roads with 20mph and 30mph limits. The disclosed PSNI figures recorded seven detections at 30mph or more in 20mph limits from 1 January to 30 November 2025, alongside 1,356 detections at 40mph or more in 30mph limits over the same stated period. The figures for September to December 2024 were five and 455, respectively.

The record gives the highest speed detected in a 20mph limit during the 2025 period as 45mph, and the highest detected in a 30mph limit as 75mph. For the four-month period in late 2024, it records 49mph in a 20mph limit and 95mph in a 30mph limit. These are examples of recorded excess speed within the scope of the FOI. They are not a census of everyone driving on 20mph streets.

The caveat printed in the disclosure is decisive: the figures exclude detections by the Northern Ireland Road Safety Partnership, and the latest available 2025 period ends on 30 November, despite the request asking about the full calendar year. Publishing 'just seven drivers broke 20mph limits in 2025' would be both factually and methodologically incorrect. The defensible wording is that PSNI's specified, incomplete dataset contained seven such detections over the stated 11 months. A fuller account would need the partnership's comparable detections and records of the time and locations monitored.

The fastest recorded incidents are not an average

The same FOI recorded a highest speed of 143mph on the southbound M2 at Rathbeg in its 2025 PSNI data, and 130mph on the M1 near Craigavon in the shorter 2024 period. PSNI's Road Safety Partnership statistics separately identify 123mph as the highest partnership-detected speed in 2025, involving a driver subject to a 45mph limit. The 143mph and 123mph figures are not competing versions of the same incident. They reflect different enforcing bodies and datasets.

The extreme values demonstrate that serious excess speed is recorded. They do not tell us the typical speed on the M2, the proportion of drivers travelling dangerously, or the number of collisions prevented by enforcement. A maximum is a single observation at the tail of a distribution. Even if a headline mentions the most extreme case, the analysis should make clear that decisions about where to deploy enforcement depend on far more than the highest speed ever detected.

Public records can reasonably support a targeted investigation of the conditions at a site where repeated severe speeding is recorded. That follow-up should examine traffic volume, proximity to vulnerable road users, injury collision history, enforcement frequency and changes following interventions. The FOIs themselves do not supply all those data, and there is no basis here for assigning motive or blame to a particular officer or driver beyond what was officially disclosed.

The 278,188 figure that changes the scale

Another disclosure, FOI/16002, records 278,188 speeding offences between 1 April 2022 and 31 July 2025 across Northern Ireland. Unlike the PSNI-only 20mph disclosure, this total explicitly includes both PSNI and Northern Ireland Road Safety Partnership detections. It also includes cases that resulted in different potential disposals, such as fixed penalties, education courses or prosecution referrals. That is why the scope is so much larger than some local tables.

A different PSNI FOI, FOI/15484, reported 3,312 speeding detections resulting in a PSNI fixed penalty notice in 2024, and 1,389 up to 31 May 2025. That table excluded detections that resulted in referrals for prosecution or courses. Comparing 3,312 with a combined PSNI-and-partnership multi-year total does not show that officers failed to penalise the overwhelming majority of speeding offences. The two counts cover different time windows, agencies and outcome types. Misusing them creates a false scandal that collapses under a basic definition check.

The discrepancies in apparent scale are precisely why an accessible public dashboard should specify every counting rule. A number headed 'speeding offences' could mean camera triggers, confirmed detections, notices issued, penalties paid, people prosecuted or courses completed. Each is a legitimate measure for a different purpose. None should be silently substituted for another.

A surveillance map without exposure is not a safety map

A serious evaluation of camera deployment would start with a defined stretch of road and an explicit time window. It would publish the type of enforcement, minutes or hours active, total eligible traffic passing through, count of speed detections by threshold, and relevant collision outcomes over multiple comparable years. Where possible, figures would be adjusted for traffic levels and changes in limits. This would let residents distinguish a fall in offences caused by more drivers slowing down from a fall caused by fewer camera hours.

Even then, proving a camera caused a reduction would require care. Other changes—roadworks, weather, junction design, new signage, traffic volumes and policing priorities—can affect speeds and collisions. A simple before-and-after table can suggest an association but not automatically show causation. Small numbers of serious collisions make local trend estimates particularly volatile. The best public-interest reporting would disclose the uncertainty, not conceal it behind a percentage.

The released Ravernet and Derry records show that relevant administrative data exists. What is missing from these public summaries is a consistent cross-site reporting frame. A defensible comparative study would need the same units and exposure measures for A6 Glenshane Road, Crescent Link and Ravernet Road, instead of treating one table's zero and another's hundreds as measurements of risk taken under identical conditions.

What officials should be asked next

The most useful follow-up requests concern operational measures: the number and duration of deployments at the five Derry sites; a matched list of partnership and PSNI detections; the end status of each recorded incident; and the operational status of fixed devices on the A24 over time. A site-level collision history, separated into fatal, serious and slight-injury incidents, would then allow researchers to examine whether enforcement intensity corresponds to local safety risks. Data on traffic volume would be needed to put the figures on a comparable basis.

Some records may be limited by operational sensitivity or other exemptions, and an authority may have defensible reasons not to disclose a live tactical deployment schedule. Historical, appropriately aggregated data is a different proposition and can often answer public questions without exposing immediate operations. The five FOIs do not establish which additional details are currently held or releasable. PureCar has not made a new information request or obtained a PSNI response for this piece; the questions are an evidence-led reporting agenda drawn from the disclosure log.

For drivers, the material also provides no defence to a speeding allegation. Camera-data inconsistencies do not change the speed limit. And a site recording zero detections in one disclosure is not a licence to assume enforcement will never occur there. The subject of the investigation is the accountability of public safety information, not a guide to avoiding cameras.

What the records prove

The five-site disclosure demonstrates a substantial spread in recorded enforcement outcomes, including 2,442 detections at one named A6 location and zero at Crescent Link. The Ravernet answer adds 106 deployments and 463 detected offences. A separate A24 response confirms that one fixed camera was not operational at the time asked about, and the 2026 20mph data comes with explicit exclusions that change how its headline figures must be read. Those are verified public records, not claims taken from a camera-warning social media group.

The original finding is that the same underlying issue—dangerous speed—is being described through several incompatible measurements. The public can see detections, financial outcomes, maximum speeds and some deployments, but not a complete, standardised picture of risk at each location. A genuinely accountable road-safety programme should make it possible to examine whether monitoring and engineering reduce risk. Until then, dramatic numbers will continue to tell stories about enforcement activity that can easily be mistaken for stories about road safety itself.

Documentary basis and limitations

This article compares PSNI FOI responses FOI-13942 (Derry-area site detections and paid fines), FOI/16341 (Ravernet deployments), FOI/00013618 (the A24 fixed camera), FOI/16823 (speeding threshold records, excluding the partnership), FOI/16002 (combined detections from April 2022 to July 2025), FOI/15484 (PSNI fixed penalties by district), and the official Road Safety Partnership statistics page. PureCar calculated illustrative ratios but has not constructed vehicle-exposure rates or claimed to have reviewed original offence case files. Historical FOI facts are anchored to their disclosed dates, not presented as real-time equipment status.