Insights

Analysing driving behaviour: a look at INVERS Driving Analysis

Speed, harsh braking, hard cornering: what the INVERS analysis measures, what operators do with it – and what data protection requires.

Roland Kainbacher – Porträt Roland Kainbacher 7 min
INVERS – Titelbild

Most damage in a sharing fleet goes back to a small number of trips. The problem: as an operator you only see the result – the scratch, the worn brake, the garage invoice. What happened during the trip stays a black box. INVERS Driving Analysis aims to close exactly that gap.

What is measured

The analysis is based on sensor data from the CloudBoxx, the telematics unit already installed in the vehicle. No additional hardware is required. Four types of events are detected:

  • Speeding – compared against the limit actually in force at that location
  • Harsh acceleration and braking
  • Hard cornering
  • Impacts and crashes

Every event is recorded with time, location and vehicle data, so what happened can be reconstructed afterwards.

From event to rating

INVERS turns individual events into two figures:

The Driving Score rates each individual trip and breaks down how it was reached. The Trip Rating translates that value into a grade from "excellent" to "critical" – meant for quick triage when hundreds of trips come in daily.

Both values can be pulled into your own systems through the API. That turns an observation into something you can act on.

What operators do with it

  • Warn early. Repeat offenders receive a message – often that is enough before anything happens.
  • Tier access. Release certain vehicle classes only after a period without incidents.
  • Adjust deposits. Higher security for conspicuous driving, lower for unremarkable driving.
  • Suspend accounts temporarily before damage occurs.
  • Reward careful driving – with credit or better rates.

The delicate part: data protection

Driving behaviour is personal data. Measuring how someone accelerates, brakes and how fast they drive means processing quite telling information about an identifiable person. That is permissible, but not unconditionally. What you as an operator need to settle:

  • Legal basis. Legitimate interest in protecting property is a sound argument, but it requires a documented balancing test.
  • Transparency. It must be clear before the first trip that measurement happens, what is measured and what follows from it. Not in the small print, but visibly.
  • Purpose limitation. Data collected to prevent damage is not a basis for other analyses.
  • Retention. Define how long events are kept, and delete afterwards.
  • Automated decisions. If a suspension happens automatically, Article 22 GDPR applies. Provide a way for a human to review the decision.

In corporate fleets, employee representation comes on top: analysing the driving behaviour of employees is generally not enforceable without involving the works council.

Note: These points are for guidance only and do not constitute legal advice. Responsibility for data protection lies with the operator. INVERS and Zemtu provide information and support, but no legal advice.

[Editorial note: to be added is whether and how Driving Analysis is integrated into Zemtu – are scores shown in the dashboard, can rules be attached to them, or is this an outlook? The figures cited come from INVERS and should be marked as their claims or confirmed.]