Fleet Fuel Theft South Africa and Why Camera Evidence Makes Your Monitoring System Work

June 2, 2026
Man siphoning diesel from a parked truck at night while a monitoring system flags the fuel drop and records the cab camera

The Alert Fires. Now You Have a Problem.

It happens at 02:30 on a Tuesday. Your fuel monitoring system sends an alert. A significant unexplained drop in tank level on vehicle fourteen. The vehicle is stationary at the overnight yard. No ignition event is logged.

You know fuel is gone. What you do not have is fuel theft evidence: who took it, how they accessed the tank, or whether this is the first time it has happened.

By the time anyone investigates in the morning, the person responsible has had hours to leave the site, dispose of the evidence, and prepare a denial. Without an independent record of what happened at that exact moment, your investigation starts and ends in the same place, with a number on a graph and a driver who says they know nothing about it.

This is the gap that fuel monitoring alone cannot close.

What a Sensor Can and Cannot Tell You

A fuel level sensor does its job correctly. It measures the volume of diesel in a tank in real time, flags anomalies, and timestamps the events. This is genuinely valuable information. It tells you that a loss occurred, approximately how much was taken, and when the level dropped.

What it cannot do is identify who was at the vehicle, what method was used, or whether the loss was theft, a mechanical leak, or an unauthorised but internally sanctioned top-up. These distinctions matter enormously.

A leak and a siphoning event can produce identical fuel level signatures. The difference between the two is not in the data. It is in the footage. Discipline a driver for siphoning when the cause was a faulty valve and you have an unfair dismissal claim. Attribute a deliberate theft to a mechanical fault and the person responsible walks away and does it again.

Partial information does not just fail to solve the problem. In some cases it actively makes it worse.

The Scale of What Is at Stake

South Africa loses approximately R3.6 billion per year to illicit fuel operations according to SARS. For fleet operators specifically, diesel already accounts for 35 to 45% of total operating costs, making it the largest controllable expense in the business. Losses that accumulate through theft or undetected leaks compound directly against the bottom line, and because most events happen overnight or in unsupervised conditions, they are rarely discovered until a reconciliation reveals a pattern that has been running for weeks.

The financial case for closing the gap between a fuel alert and actionable evidence is straightforward. Every event that cannot be investigated is an event that will likely repeat.

What Video Evidence Changes

When a fuel monitoring alert is paired with a camera system covering the same vehicle and timestamp, the investigation that was previously impossible becomes immediate and conclusive.

Siphoning produces a specific visual signature. Someone approaches the vehicle during the stationary window, accesses the tank, and leaves. The footage shows who it was, what equipment they used, and how long the event took. The combination of the fuel level graph and the footage is evidence that can support disciplinary proceedings or a criminal complaint to SAPS.

Unauthorised refuelling produces a different signature. The tank level rises unexpectedly. The footage shows whether a legitimate refuel took place outside of authorised hours, or whether fuel was added to conceal an earlier siphoning event by topping the tank back up partially.

Leaks and mechanical faults look different again. A leak produces a gradual decline rather than a sharp step-down. The camera confirms no human involvement during the period of loss. This clears the driver, identifies a maintenance issue, and removes any ambiguity from the record.

In each case, the sensor provides the trigger and the timestamp. The camera provides the context that makes the trigger meaningful.

The Evidence Standard That Makes Dismissal Stick

A driver is accountable for the fuel under their care. If fuel disappears during a shift and they cannot explain it, that is a disciplinary matter. Most fleet operators know this. The problem is not willingness to act. It is having evidence strong enough that the action holds.

A dismissal for theft challenged at the CCMA requires more than a fuel level graph. A driver who denies involvement can point to the same data and argue a leak, a faulty sensor, or a draw that occurred before or after their shift. Without an independent visual record, that denial is difficult to disprove. Employers who dismiss on the strength of sensor data alone face the risk of a contested hearing with no footage to support their case.

Video footage tied to the same timestamp as a fuel alert closes that gap. It shows the person, the act, and the time. Combined with the fuel monitoring record, it creates a complete chain of evidence that is defensible at a disciplinary hearing and admissible in support of a criminal complaint.

Without video, the investigation ends at "we believe someone stole fuel." With video, it ends at "this person stole fuel at this time using this method, and here is the footage."

How the Integration Works

Digisky's live fuel monitoring connects directly with the AI fleet camera system. When a fuel drop triggers an alert, the platform allows you to pull the corresponding timestamp footage from the vehicle camera immediately. You do not need to request a manual review or wait for a technician to retrieve a card. The footage is available remotely, in real time, from any device.

This matters most in the hours immediately after an event, when the footage is freshest, the evidence trail has not been disturbed, and any follow-up action such as reviewing yard access logs or identifying vehicles present at the site can still be taken while the information is current.

The system identifies the cause, matches it to a driver and location, and produces the documentation needed to act. A sensor alert without this integration is a notification of a problem. A sensor alert with camera integration is the beginning of a resolution.

For fleet operators who want to understand what the camera system captures and how it integrates with fuel monitoring, the Digisky AI fleet camera system provides the dual-facing, always-on video coverage that makes this level of investigation possible.


Digisky's live fuel monitoring and AI fleet camera systems work together to give you financial data and visual proof in the same platform. When an alert fires, you can see exactly what happened, who was responsible, and what evidence you have to act on it. Contact us to discuss integrating fuel monitoring and camera coverage across your fleet.


Share: