Geofence and location data in injunction and bail applications

Injunction proceedings with exclusion zones, non-molestation orders with location conditions, bail conditions requiring a defendant to remain at or away from an address: all of these now generate a common category of evidence, the location record. It arrives from several directions. Phone tower and cell-site data from telecoms providers, location history stored in Google and Apple accounts, geofence alerts generated by monitoring apps installed as a condition of bail, telemetry from tags, and the location metadata embedded in photos and messages the subject has sent voluntarily.

Each of these sources has a different accuracy profile, a different chain of custody, and a different failure mode. Treating them as interchangeable is the most common error in both directions: respondents are accused on the basis of data that cannot support the accusation, and genuine breaches go unchallenged because the opposing party does not understand what to ask for.

The sources and what they are actually worth

Cell-site data. Call detail records from a network operator map a handset to the cell that carried each call or message, with the cell's coverage footprint. At its best this places a phone in a general area at a specific time. It does not place a phone at an address. Cell coverage in dense urban areas may span hundreds of metres and overlap between cells; a handset can register to a cell it is not closest to. Cell-site evidence supports "in the vicinity" conclusions and exclusions, not pinpoint positions. It is strongest as rebuttal: if the phone was connected to a cell two kilometres from the exclusion zone at the alleged time, that is genuinely exculpatory.

Google and Apple location history. Where enabled, these produce timestamped position fixes of far higher quality, often derived from GPS with Wi-Fi assistance, recorded continuously or at event triggers. For civil injunction proceedings this is frequently the richest source available, because it can be lawfully obtained through disclosure from the opposing party's own device or account, rather than from an operator. But it is app-managed data: it can be paused, deleted in part, or affected by the account being logged in on another device. The examiner's job includes establishing which device generated each fix, which is a more involved question than it first appears and overlaps with the issues we discuss in our article on chat exports as evidence.

Tag and monitoring app telemetry. Court-ordered location monitoring, whether a physical tag or a smartphone app installed as a bail condition, produces a continuous record designed for exactly this evidential purpose. The critical questions are operational rather than technical: the calibration of the exclusion polygon in the monitoring system, the handling of GPS drift near zone boundaries, the alert review process, and whether the log records raw fixes or only triggered alerts. A log that shows only alerts hides the near-misses, and near-misses are how you assess whether an alert reflects genuine intrusion or normal positional noise.

Embedded location in media and messages. Photos with GPS EXIF, and messaging platforms that attach location, provide point-in-time fixes that are often volunteered by the subject. Their value is documentary rather than continuous: a single photograph taken inside an exclusion zone at a prohibited time is a discrete, verifiable fact. Our treatment of the metadata itself, and how it is verified and challenged, follows the approach set out in file metadata in litigation.

Accuracy, drift and the boundary problem

Exclusion zones make GPS accuracy a legal question. Consumer GPS fixes are typically accurate to a few metres under open sky, but degrade to tens of metres among tall buildings, indoors, or in vehicles. Urban canyon effects can pull a fix tens of metres in a consistent direction, which is precisely the pattern that generates false breach allegations near zone edges. A competent examination of alleged breaches plots the fixes over time and looks at the scatter. A track that runs cleanly along a road and briefly clips a zone corner by eight metres reads very differently from a track that shows a deliberate approach, dwell, and departure.

The examiner also checks the fix source. Phones blend GPS, Wi-Fi positioning and cell triangulation, and the recorded accuracy field, where present, is evidence. A fix with a stated accuracy radius of 50 metres cannot, on its own, establish entry into a 100-metre exclusion zone with any confidence at all.

Whose phone was it

Location data locates a device, not a person. Attribution requires corroboration: the device was in the subject's possession, the account is theirs, the phone was not left at home or handed to someone else. Opposing parties routinely assert attribution as though it were automatic. It is not, and a location record unaccompanied by any attribution evidence is vulnerable precisely where it matters most. The same principle, that a device record is not a person's record until proven so, runs through all handset-based evidence, as we noted in the vehicle telematics context.

How the evidence is prepared for court

• Acquisition with provenance. Data is obtained from the device or account by a method that preserves the original: a forensic extraction of the handset, a properly scoped production of the account history, or the operator's records under the applicable disclosure route. Screenshots of a map UI are exhibits, not evidence, because they show a rendering rather than the record and cannot be re-examined.
• Verification. The examiner checks internal consistency: clock sources and timezone handling, gaps in the record, signs of deletion, and agreement between independent sources. Where Google history and cell-site data tell the same story through different systems, the corroboration is real evidence, not decoration.
• Presentation. Plotted tracks with the exclusion zone overlaid, a chronology of entries and exits with accuracy context, and a stated methodology so the opposing expert can test the work rather than merely disagree with it. Under CPR Part 35 duties in civil proceedings, that methodology and the material relied upon must be disclosed.

What location evidence does not prove

It does not prove intent, and courts know this. It does not prove identity on its own. It cannot retroactively create a zone that the order did not clearly define, and disputes over ambiguous zone descriptions are drafting disputes, not forensic ones. And a gap in the record proves nothing either way: a phone whose location history stops an hour before an alleged incident is suspicious, incomplete, or simply off, and the examiner's job is to say which of those the evidence supports rather than to fill the silence.

For instructing solicitors, the practical sequence is: preserve early, before histories are deleted; specify the device and account in disclosure; demand the accuracy fields and not just the plotted positions; and instruct an examiner who will state the limits of the evidence as clearly as its strengths. The location record is often decisive. It is decisive most reliably in the cases where nobody pretended it was more than it is.