Guide
What is a Geofence?
Drawing a fence takes a minute. Getting it to fire at the right moment depends on how often the tracker reports, how accurate each point is, and what the software does when a point lands on the line. This guide covers those parts.
- A geofence is evaluated in software after a position arrives. A tracker that does not report cannot trip one.
- Reporting interval and GNSS error decide whether a fence works. Size and debounce fences for both.
- Enter, exit, dwell, and “has not reported” are different events with different uses.
- TruckerPro lists geofences as planned pilot work, not a live feature.
How a geofence is evaluated
The application stores a geometry: a point plus a radius, or a polygon. Each accepted location report is tested against it. Is this point inside or outside, and was the previous point on the other side? That change of side is what becomes an enter or an exit. Dwell is “inside for longer than N minutes.” None of this runs on the satellite; it runs in software after the point arrives.
For a radius fence the test is a distance calculation from the centre. For a polygon it is a point-in-polygon test, commonly done by casting a line from the point and counting how many edges it crosses. Both are cheap to compute. The hard part is not the maths; it is the quality and spacing of the points being tested.
Some trackers can also hold simple zones on the device itself and report the crossing as an event. Whether a particular model does, how many zones it holds, and which shapes it accepts are in that model’s configuration documentation. TruckerPro has not validated on-device zones on either pilot candidate.
The four events, and what each is for
| Event | What triggers it | Typical use |
|---|---|---|
| Enter | First accepted point inside after one or more points outside | Arrival at a customer, return to the yard |
| Exit | First accepted point outside after points inside | Departure timestamps, a trailer leaving a lot after hours |
| Dwell | Still inside after a set time | Detention evidence, equipment idle on a jobsite |
| Has not reported | No point at all for longer than expected | A dead battery, a damaged unit, or one taken out of coverage |
The first three need positions. The fourth needs the absence of them, which is why it is the one most often forgotten and the one that catches the problems the others miss.
What fleets use them for
- Yard control: a trailer left after hours.
- Customer sites: arrival and departure timestamps for detention.
- Jobsites: equipment that should not leave the fence.
- Corridors: a unit that wandered off an expected route—only if reports are frequent enough.
For how arrival and departure events feed dispatch, customer updates, and invoicing inside a TMS, see how a TMS uses GPS data.
Reporting interval decides what you can catch
A fence can only see the points it is given. A truck at 100 km/h covers about 1.7 km per minute, so a tracker reporting once a minute leaves gaps that long between points. A highway-side fence smaller than that gap can be driven straight through without a single point landing inside it. Dwell has the same limit: a battery tracker that wakes every few hours cannot show that a trailer sat at a dock for ninety minutes.
This is a reporting-profile problem, not a “geofence bug.” Hardwired units can usually afford frequent reports; battery units trade report frequency for runtime. The hardwired vs battery guide covers that trade. Decide which event matters for each asset class, then choose the interval that can detect it.
Size the fence to position error
GPS.gov states that GPS-enabled smartphones are typically accurate to within a 4.9 m (16 ft) radius under open sky, and that accuracy worsens near buildings, bridges, and trees. Dedicated trackers face the same physics. A fence drawn tight around a dock door beside a steel warehouse will see points scatter across its edge while the trailer has not moved.
Two habits keep this under control. First, draw fences larger than the error you expect at that site, with the most margin where the sky view is blocked. Second, debounce: require two consecutive points on the new side, or a minimum time inside, before raising the event. That one rule removes most duplicate enter and exit alerts.
Design rules that keep geofences useful
- Name fences after places people recognize (“Calgary drop lot”), not “Fence 17.”
- Use a polygon where neighbouring sites sit close together, so one fence does not swallow the next property.
- Size them to GNSS error. A 20-metre circle around a steel warehouse will chatter.
- Alert a person who can act. A mailbox nobody reads is not a control.
- Keep a silence alarm. Stolen or dead units often stop reporting rather than politely exiting.
- Review fences when sites change. A customer that moved docks or a yard that expanded makes old fences wrong without anyone noticing.
Privacy and notice
A fence around a yard or a customer is an operational tool. A fence around a driver’s home, or one that stays active outside working hours on a vehicle used personally, records much more than the business needs and can create legal exposure. Decide in writing which fences exist, who sees their events, and how long those events are kept. The legal notice guide covers the general questions to take to counsel.
Questions to ask a vendor about geofencing
- Are fences evaluated on the device, on the server, or both, and what happens to a crossing that occurs while the device is out of coverage?
- Which shapes are supported, and is there a limit on fences per account or per device?
- Is there debounce or a minimum dwell before an event is raised, and can it be set per fence?
- Is “has not reported” a separate alert, and can its threshold follow each asset’s reporting profile?
- Are event timestamps the device’s position time or the server’s receipt time, and are both exported?
Motive and Samsara both productize geofences on top of high-frequency vehicle pings. A future TruckerPro application would need the same building blocks: trusted points, stored geometry, debounce, and delivery. Those blocks are not shipped.
Source
- GPS.gov, GPS Accuracy (U.S. government): smartphone accuracy under open sky and the effect of buildings, bridges, and trees.
Related reading
- Glossary: geofence
- How a TMS uses GPS data
- Hardwired vs battery GPS tracker
- Is GPS tracking company vehicles legal?
- How trailer GPS tracking works
Questions
What is a geofence in fleet tracking?
A virtual perimeter around a yard, customer, jobsite, or corridor. The application compares reported coordinates to that shape and can raise an event.
Does a geofence work if the tracker is offline?
No. If the device cannot send a point, the fence cannot evaluate it. “Has not reported” is the event you want for silence.
How big should a geofence be?
Larger than the position error at that site plus the distance an asset can travel between reports. Open yards tolerate tighter fences than sites beside tall buildings, bridges, or trees, where GPS accuracy worsens.
Why did a geofence fire twice when the truck only arrived once?
Usually because consecutive points scattered across the boundary. Requiring two consecutive points on the new side, or a short time inside, before raising the event removes most of this chatter.
Can I set geofences in TruckerPro GPS today?
No. Geofence alerts are a planned capability. They are not offered as a production service.
Share a pilot use case
Tell us what you would track and where it operates. This records interest; it does not place an order.