EEG in Race Cars: Why Measuring the Driver's Brain Is So Hard

Why the cockpit is a hostile environment for EEG

Electroencephalography records the brain's electrical activity through the scalp. The signal is very weak, and anything that generates electricity or moves the electrodes disturbs it. These disturbances are called artefacts.

A race car piles them up. Researchers who have recorded EEG in a real car describe it as a hostile recording environment ^[1].

• Head movements, at every braking zone and every corner.
• Activity in the neck and jaw muscles, which produces a far stronger signal than the brain does.
• Eye movements and blinks, which mainly contaminate the frontal electrodes.
• Vibration transmitted through the seat, the chassis and the helmet.
• Heat and sweat, which alter the contact between electrode and skin.

What studies in real driving show

In 2018, a research team recorded the EEG of drivers in a real car travelling at 40 km/h on a closed track, using 64 electrodes. Classic brain signatures, such as the P300 response and theta and alpha patterns, were comparable to those recorded in the lab ^[1]. Measurement is therefore possible.

But quality comes at a cost: 28.89% of trials had to be discarded in real driving, against 14.11% in the lab ^[1]. And those conditions remain far from a race car, in speed, in accelerations and in vibration.

Most work is therefore carried out in simulators. Even there, researchers stay cautious: in a recent study, the authors could not rule out that eye movements had contaminated frontal theta ^[4], which happens to be one of the markers of cognitive load.

Motion artefacts: methods and their limits

A 2022 review assessed the methods used to clean movement-induced artefacts. Independent component analysis, which mathematically separates the sources of the signal, is the most widely used ^[2].

The review makes one point clearly: there is no universal method. The right processing depends on the type and intensity of the movement, and the authors call for new approaches to be developed ^[2]. Cleaning the signal afterwards is no substitute for a clean recording at the source.

Dry or wet electrodes?

Wet electrodes, with conductive gel, offer the best contact, but they take a long time to fit and the gel dries out. Dry electrodes go on in seconds, which makes them better suited to the field.

A comparison in 32 people at rest found no significant difference in alpha or beta power between the two systems. The dry electrodes did, however, record more activity in the low frequencies, theta and delta ^[3].

That detail matters: theta is a key marker of cognitive load. Equivalence shown at rest does not automatically carry over to a driver in motion.

Synchronising the brain and the car

A brain signal is only useful to an engineer if it is placed at the right point on the track. EEG, cardiac signals and telemetry therefore have to be aligned on a single clock.

Research tools such as Lab Streaming Layer synchronise heterogeneous streams to the millisecond through timestamping. Their authors point out, however, that these tools do not measure the internal delays specific to each device ^[5].

At 200 km/h (about 124 mph), a car covers roughly 55 metres per second. An offset of a few tens of milliseconds places the measurement several metres away: before or after the braking point. Timing precision is not a technical detail, it is the condition for the data to be usable.

What this means for measurement in racing

Measuring a driver's brain activity in the car means dealing with each difficulty at the source: an instrument engineered for vibration, G-load, heat and motion, a check of signal quality before going out on track, and strict synchronisation with the car.

That is the position SPARK MOTORSPORT takes: starting from the constraints of the cockpit, where most brain-sensing setups stop at the simulator or the lab.