Geofencing and Fleet Alerts Built Against Alert Fatigue
An alert that fires forty times a day is not a control. It is noise with a notification sound. KO Fleetz lets you draw the zones that matter, attach rules with dwell times and schedules, and route each event to the one person who can actually do something about it.
Everyone turned the notifications off in the second week
It starts well. Someone draws a circle around the depot, switches on entry and exit alerts, and for two days the operations group chat is genuinely useful. Then the circle turns out to clip the main road. Every vehicle passing the depot at forty miles an hour triggers an arrival and a departure inside the same minute.
By the end of the fortnight the channel has produced several hundred messages, of which perhaps four mattered. People stop reading. Someone mutes it. The alert that should have caught a trailer leaving a secure yard at two in the morning arrives in a feed nobody has opened since Thursday, and it sits there next to a thousand cars driving past a depot they never entered.
The other failure is quieter. Alerts get sent to a distribution list rather than a person, so the responsibility spreads out until it disappears. A notification everybody receives is a notification nobody owns, and the one at 03:00 gets read at 09:00 by which time the trailer is somewhere else entirely.
Fewer alerts, each one worth reading
KO Fleetz draws zones as polygons that follow the actual boundary, not circles that approximate it. A depot with a road down one side gets a shape that excludes the road. That single detail removes most nuisance triggers before any clever filtering is needed, because the geometry stopped lying about where the site is.
Rules then add the conditions that separate an event from a coincidence. Entry, exit and dwell beyond a threshold. Schedules, so a yard that is supposed to be busy at eleven in the morning and empty at midnight can hold different rules for each. Vehicle scope, so a rule applies to the tankers rather than to everything with a tracker.
Routing is where alerts stop being a group chat. Each rule names who is notified, by which channel, and what happens if nobody acknowledges. Overnight events reach the person actually on duty overnight. KO Fleetz also declines to alert on things it cannot substantiate: a brief position wobble at a zone boundary is treated as GPS noise, not as a vehicle leaving and re-entering nine times.
Capabilities
What KO Fleetz geofencing & fleet alerts gives your team
Polygon zones that follow the fence line
Draw the real boundary of a site rather than dropping a radius over it, so the public road running alongside your depot stops generating arrivals.
Dwell-time thresholds
Alert only when a vehicle stays inside a zone longer than it should, which turns a stream of entry pings into the one stop that has gone wrong.
Time-of-day rule schedules
The same yard can be unremarkable at noon and alarming at midnight. Rules carry their own hours instead of applying identically around the clock.
Boundary noise suppression
A vehicle parked on a zone edge drifts across it as fixes scatter. Repeated crossings in a short window read as GPS jitter, not nine departures.
Named ownership per rule
KO Fleetz names one responsible person on every alert, by channel, with an escalation path if unacknowledged. Nothing lands in a group where everyone assumes someone else has it.
Vehicle and zone scoping
Apply a rule to a class, a depot, a contract or a single trailer, so a restriction that matters for tankers does not fire for every van in the fleet.
Unauthorised area alerts
Flag entry into zones a vehicle has no business being in, including restricted areas, competitor sites and the exclusion zones your insurer asked about.
Site arrival and departure feed
Clean arrival and departure timestamps per site become the input for turnaround measurement and customer waiting-time conversations.
How it works
How KO Fleetz does it
Step 1: Draw the zones that carry consequences
Start with the sites where something happens if a vehicle arrives, leaves or lingers. Mapping every location you have ever visited produces a map, not a control.
Step 2: Attach a rule and a threshold
Decide what condition is genuinely worth interrupting someone for, and what dwell time separates a normal stop from one that needs a phone call.
Step 3: Name the person, not the group
Each rule gets an owner, a channel and an escalation path. An unacknowledged alert has to go somewhere other than back to itself.
Step 4: Prune what nobody acts on
Review which rules fire and which ever produce action. A rule that has raised two hundred events and caused zero interventions should be deleted, not tolerated.
Outcomes
What changes
- Alerts that survive the second week
- Read, not muted
- Polygons instead of approximate circles
- Real boundaries
- Per rule, with an escalation path
- One owner
- Events reach whoever is actually on duty
- Overnight
Frequently asked questions
Three fixes, in order of impact. Draw polygons that match the real site boundary, because most nuisance triggers come from a radius that overlaps a road. Add dwell thresholds so passing through does not count as arriving. Then delete rules that have never once produced an action. The instinct is to filter harder, but the usual culprit is geometry, and no amount of filtering rescues a badly drawn zone.
Bigger than your position error, which in practice means a zone tighter than roughly fifty metres will misbehave. GPS scatters by a few metres under open sky and considerably more beside tall buildings, so a small zone will register crossings that never physically happened. If you need to know which bay a vehicle is in rather than which site, geofencing is the wrong instrument.
No, and this is worth being clear about. A dwell alert tells you a truck has been at a customer for ninety minutes. It cannot distinguish waiting for a bay from a long unload from a driver taking a break. What it does is put the question in front of someone while the vehicle is still there and there is still time to ring. Interpretation stays human.
Buffered fixes arrive late, so the entry and exit reconstruct correctly but the alert does not fire on time. This matters for covered terminals and underground loading areas, where signal reliably drops. The honest position is that geofencing degrades to a delayed record in those places. If a site regularly swallows signal, plan around it rather than expecting real-time alerts to work there.
The technical limit is not usually what constrains you. The practical limit is how many rules a team can maintain without them decaying into noise. Fleets do better with thirty zones that each carry a consequence than with four hundred that map every place a van once parked. Zones are cheap. Attention is not, and attention is the resource the whole system is spending.
That is arguably their more valuable use. The clean arrival and departure timestamps KO Fleetz records per site are the raw material for turnaround and detention measurement, and they beat driver-reported times because nobody has to remember to record them. The same events can trigger job status changes and customer notifications. Treat the alert as one output of the zone rather than its whole purpose.
Draw a zone that earns its notification
Point KO Fleetz at the site that generates your worst alert noise, and get back a rule people will still be reading next month.