Seeing Machines challenges BAC with in-vehicle tech

Australian driver monitoring specialist Seeing Machines has released the first in a new series of technical papers on non-fatigue driver impairment, focusing on alcohol and questioning the reliance on Blood Alcohol Concentration as the primary measure of risk.

The company said the research shows that "alcohol-related impairment does not increase or decrease in a simple linear relationship with BAC." Driver performance can deteriorate or recover at different stages during intoxication, even when roadside tests record similar or falling BAC levels.

The paper argues that this pattern creates potential blind spots for enforcement strategies that depend on single-point roadside tests. It says that in-vehicle driver monitoring can track how impairment actually affects behaviour while a person is behind the wheel.

The work forms part of a broader research programme on non-fatigue impairment that will include cannabis and other substances in future instalments.

Limitations of BAC

Seeing Machines said its latest research highlights the limits of using BAC alone as a benchmark for impairment. It said traditional approaches often miss how impairment unfolds over time as alcohol is absorbed and metabolised, and how this affects real-world driving tasks.

The company's driver monitoring systems use vision-based technology that tracks a driver's behaviour inside the vehicle. The paper says this approach allows direct measurement of functional impairment in the cabin as it develops.

The company argues that this method can support earlier identification of risky driving and intervention while the journey is in progress.

The system observes indicators such as attention, gaze, and control inputs, rather than inferring risk only from chemical tests.

Alcohol remains a significant contributor to road fatalities and serious injuries worldwide. Many jurisdictions have introduced extensive roadside testing programmes, but the toll from alcohol-related crashes persists.

"Roadside alcohol testing has saved countless lives, but it doesn't tell the whole story. In-vehicle systems allow us to detect how impairment actually affects a driver in real time - not just what a single BAC reading says," said Dr Mike Lenné, Chief Safety Officer, Seeing Machines.

Temporal dynamics

The new paper focuses on the changing nature of impairment during both the ascending and descending phases of intoxication. The company said "impairment can worsen as BAC rises but can also remain significant, or fluctuate, as BAC falls."

This temporal effect can cause a mismatch between measured BAC and the functional ability of a driver. A person with a falling BAC may still show reaction time delays, poor judgement or reduced situational awareness.

Seeing Machines said its driver monitoring technology observes impairment as it happens. The system does not depend on knowledge of the driver's BAC level.

In some cases, the company said performance can deteriorate even when BAC is relatively low, or can briefly stabilise despite a high BAC, due to individual tolerance and other factors. This variability can make regulatory thresholds a blunt instrument for assessing moment-to-moment risk.

"Roadside alcohol testing has delivered great reductions in road injury over many years. Our goal is to complement these existing roadside programs by leading the introduction of in-vehicle approaches," said Lenné. "By incorporating the temporal dynamics of impairment, how it evolves during both the ascending and descending phases of intoxication from alcohol as well as the complex effects of other drugs such as cannabis, our technology can reflect real-time real-world risks and provide greater safety protections to road users."

Academic partnerships

The technical paper draws on collaborative research with universities and impairment experts. Seeing Machines said it is conducting experiments that examine the disconnect between BAC readings and performance measures such as lane-keeping, reaction times and error rates.

The company said these studies use controlled conditions to compare breath or blood alcohol levels against observed driving behaviour and in-vehicle monitoring outputs. It said findings from this work are feeding into product development and safety frameworks.

Seeing Machines said the collaborations support its focus on aligning monitoring technologies with real-world safety outcomes on public roads and in commercial fleets.

The first paper in the series is now available to industry stakeholders and researchers.

Future editions will examine impairment linked to cannabis use and other non-fatigue factors, and their detectability by in-vehicle systems.

The company said the wider research series will look at how driver monitoring can address a range of impairment sources beyond tiredness, as regulators and manufacturers consider new safety standards for vehicles worldwide.