Intervening before the crash
Indicators such as fatigue, lane departure and tailgating turn into warnings before a crash; the driver is alerted inside the cab immediately.
Fleet · Driving Safety
Driver safety scoring and behaviour monitoring
Driving behaviour that is not measured cannot be managed. Our scoring module collects nineteen parameters from telemetry and cabin camera into a single risk index, produces a success percentage for every driver and ranks them within the fleet.
Scoring is normalised by distance. Comparing a driver who covers eight hundred kilometres with one who covers two hundred and fifty on raw violation counts would be unfair; the measure is therefore violation density per kilometre, not violation count. A good score cannot be earned by driving less.
- Risk Index
- Success Percentage
- Speeding
- Over-revving
- Harsh Cornering
- Lane Departure
- Tailgating
- Idling
- Fatigue
- Seatbelt
- Mobile Phone
- Fuel Ratio
01 WHY MEASURE
Driving safety:
Driving safety:
behaviour that is not measured cannot be managed
A report filed after a crash is a record; behaviour measured before the crash is an opportunity.
In heavy vehicle fleets safety is usually discussed after the event: the records are revisited when a crash, damage, fine or customer complaint arrives. The limit of that approach is obvious — the event has already happened. Measuring driving behaviour continuously makes the same information available before the event.
Measurement pays off in three places. On the safety side the components of crash risk become visible, on the cost side the fuel and maintenance bill of aggressive driving is separated out, and on the management side driver evaluation rests on data rather than opinion.
Fuel and maintenance
Harsh acceleration, high engine revolutions and long idling directly increase fuel consumption and wear; these items can be attributed to individual drivers.
Objective evaluation
Driver evaluation rests on measurement rather than opinion; bonus, training and assignment decisions gain a defensible basis.
02 PARAMETERS
Nineteen parameters read from
Nineteen parameters read from
telemetry and camera
Part of behaviour is visible in vehicle data, part only in cabin and road footage.
| Parameter | Source | What it indicates |
|---|---|---|
| Speeding | Telemetry | How often the road or fleet speed limit is exceeded |
| Over-revving | Telemetry · CAN-bus | Running the engine in an inefficient rev band |
| Harsh braking | Telemetry | Lack of following distance and anticipatory driving |
| Harsh acceleration | Telemetry | Driving that increases fuel consumption and wear |
| Harsh cornering | Telemetry | Entering a bend at speed; risk of load shift |
| Idling violation | Telemetry | Fuel and engine hours burned while stationary |
| Aggressive driving | Telemetry | A driving pattern in which several violations occur together |
| Lane departure | Camera · ADAS | Leaving the lane without signalling |
| Tailgating | Camera · ADAS | Closing below the safe following distance to the vehicle ahead |
| Fatigue | Cabin camera | Eyes closing, head dropping, prolonged gaze deviation |
| Mobile phone | Cabin camera | Phone use while the vehicle is moving |
| Seatbelt | Cabin camera | Belt not fastened while the vehicle is in motion |
| Smoking | Cabin camera | Fire risk in dangerous goods transport |
| Distance covered | Telemetry | The base value the score is normalised against |
| Fuel consumed | Telemetry · CAN-bus | Real consumption read from the vehicle |
| Fuel ratio | Calculated | Litres consumed per 100 kilometres |
In vehicles without a cabin camera the last group of parameters is not measured; the score is calculated from telemetry-sourced violations alone. Coverage changes with vehicle hardware, the measure stays the same.
03 SCORING MODULE
Score, ranking and
Score, ranking and
period comparison
A module inside the platform where drivers are scored and managed.
The module reduces measured violations to two indicators. The risk index is a raw weighted counter: the total, weighted by violation type, is normalised by distance. The success percentage shows the driver's position within the fleet; the best driver approaches 100, the riskiest falls to 0.
The module works by period. The same driver's comparison with the previous period, their position against the fleet average and the violation type they concentrate in all appear on one screen. Ranking is within the fleet; drivers from different fleets are not compared with each other. In the Turkish product line this module is named Sürücü Ligi, the Driver League.
| Driver | Success | Risk index | Distance | Speeding | Over-revving | Harsh cornering | Lane departure |
|---|---|---|---|---|---|---|---|
| Driver A | 100% | 131 | 833 km | 42 | 0 | 4 | 0 |
| Driver B | 98% | 161 | 250 km | 20 | 88 | 8 | 0 |
| Driver C | 91% | 251 | 304 km | 19 | 156 | 6 | 26 |
| Driver D | 60% | 664 | 737 km | 81 | 348 | 30 | 202 |
| Driver E | 0% | 1477 | 308 km | 483 | 136 | 280 | 10 |
The values in the table come from a real period report with driver identities anonymised. Driver E is the one to note: 483 speeding and 280 harsh cornering violations over 308 kilometres push the risk index to 1477 and zero the success percentage. Driver A, by contrast, covered the longest distance yet holds the lowest index.
The comparison between Driver B and Driver C is instructive too. They cover similar distances with similar speeding counts, but C has almost twice the over-revving and 26 lane departures on top. That difference shows in the score and states directly what the training subject should be.
04 FAIR SCORING
A good score cannot be earned
A good score cannot be earned
by driving less
A system that ranks on raw violation counts punishes the hardest-working driver.
This is the most common flaw in driver scoring. When violations are totalled as raw numbers, the long-haul driver inevitably accumulates more of them and drops to the bottom of the list. The result is that the system loses its credibility in the eyes of the drivers it is meant to guide.
The risk index is therefore normalised by distance. The measure is not violation count but violation density per kilometre. In the table Driver A holds an index of 131 with 58 violations over 833 kilometres, while Driver E holds 1477 with 916 violations over 308 kilometres; the gap comes from density, not from distance.
Violations are weighted by type
Not every violation type carries the same weight; fatigue or tailgating and an idling violation represent very different levels of risk.
Normalised by distance
The weighted total is divided by the kilometres the driver covered during the period; comparison then happens on the same scale.
Ranked within the fleet
The normalised index is ranked within the fleet and converted into a success percentage; the benchmark is the fleet's own reality.
Fuel is reported separately
Fuel ratio is not mixed into the safety score. Totalling the consumption of a heavy tractor unit and a light commercial vehicle in one score would be misleading; efficiency is tracked as a separate indicator.
05 FROM SCORE TO ACTION
A score is not a conclusion,
A score is not a conclusion,
it is a starting point
Ranking alone does not change behaviour; what changes it is what the ranking says.
- Targeted training: The dominant violation type is visible per driver; if lane departure stands out in one and over-revving in another, two separate subjects are worked on.
- Immediate feedback: When a violation is detected the driver is warned audibly and visually inside the cab; correction does not wait for the end of the period.
- Period comparison: The same driver's difference from the previous period shows whether the training actually worked.
- Bonus and recognition: The score becomes a defensible criterion in bonus and recognition processes; argument over opinion gives way to data.
- Assignment: In dangerous goods or sensitive cargo transport, prioritising low-risk drivers becomes possible.
- Evidence-based review: Because footage of the moment and telemetry data are stored together, dispute processes move faster.
Driver scoring data can also be queried in natural language through CreaAI: without opening a report screen you ask something like "which drivers increased lane departures most this month", and the answer turns into an Excel file in one step.
06 BY SECTOR
The same scoring,
The same scoring,
different weights by sector
Which violation counts as critical changes with the cargo carried and the operation run.
07 FREQUENTLY ASKED QUESTIONS
Frequently asked questions
Frequently asked questions
about driver scoring
The questions we meet most often in project discussions.
What is driver safety scoring?
It is a module running inside the Createch platform. It scores driver behaviour measured from telemetry and camera data, produces a risk index and a success percentage for every driver, and ranks drivers within the fleet on that score. The aim is not to watch the driver but to improve behaviour by making it measurable and comparable. In the Turkish product line the module is named Sürücü Ligi, the Driver League.
Which parameters is the driver score calculated from?
Telemetry supplies speeding, over-revving, harsh braking, harsh acceleration, harsh cornering, idling violations and aggressive driving. Camera and ADAS supply lane departure, tailgating, seatbelt, mobile phone use, smoking and fatigue. Distance covered, fuel consumed and fuel ratio are reported alongside them.
Does a driver who covers less distance gain an advantage?
No. The risk index is calculated from violation density normalised by distance, not from raw violation counts. A driver covering 833 kilometres and one covering 250 kilometres are compared on the same scale; showing low risk by driving less is not possible.
How do the risk index and the success percentage relate?
As the risk index rises, the success percentage falls. In a sample period report a risk index of 131 corresponded to 100 per cent success, 664 to 60 per cent and 1477 to 0 per cent. The index is a raw counter; the success percentage is a normalised indicator of position within the fleet.
Is fuel consumption included in the safety score?
No, it is reported separately. Fuel ratio varies greatly by vehicle type; totalling the consumption of a heavy tractor unit and a light commercial vehicle in the same score would be misleading. The safety score comes from behaviour violations, while fuel efficiency is tracked as a separate indicator.
How are the scores used in driver training?
A score is not a conclusion on its own but a starting point. Because the dominant violation type is visible per driver, training can be targeted rather than generic. Period comparison then shows whether the training worked.
08 CONTACT
Manage your drivers with data
We determine which parameters can be measured according to the hardware already fitted to your fleet. In vehicles without a cabin camera the score is built from telemetry; when a camera is added the coverage widens while the measure stays the same.
Get in touch for detailed information, project consultancy and demo requests. Our technical team is ready to plan the scope of the installation according to your fleet size and existing hardware.