Publication

Carbon dioxide (CO 2) is an odourless and colourless compound present in the atmosphere at a concentration of approximately 400 parts per million (or ppm). While CO 2 is not toxic per se, exposures to very high concentrations (> 10000 ppm) may have moderate to severe implications for human health, e.g., high blood pressure, dizziness, nausea or even life-threatening conditions such as hypercapnia or loss of consciousness. Short-term exposures to moderate-high levels (1000-5000 ppm) are also known to produce mild symptoms such as fatigue, discomfort or headache. More recently, studies have shown they could also affect the human physiological and cognitive state. Some of these effects include increases in brain activity related to sleepiness, lower heart rate and worse cognitive performance in tests assessing concentration or decision-making capacity, among others. These effects have been linked to performance problems in ecological environments with elevated CO2, including schools, offices and airplanes.

Moderate-high CO 2 levels have also been detected inside vehicles driving with the HVAC unit (heating, ventilation and air conditioning) in recirculating mode (RC). This system is often used to reduce the inflow of pollutants from outside (e.g., when traffic is congested) and/or maximize the air conditioning effects, especially during the summer. In e-cars, drivers also activate the RC to reduce energy consumption when cooling down the warm air from outside. A consequence of enabling the RC system, however, is a rapid build-up of respiration-derived CO 2 due to a lower air exchange rate. In fact, different studies have detected moderate-high CO 2 levels (1000 – 5000 ppm) after just a few 5-10 minutes of driving. This evidence suggests that a large number of drivers may frequently expose themselves to CO2 concentrations that could affect their state and, more importantly, their ability to drive safely. This potential effect, however, has received no attention in the literature.

To address this knowledge gap, the present project aimed to elucidate the possible effects of moderate- high CO 2 levels on driver fitness and performance. In addition, this project analyzed the influence of other factors, such as driver mental workload and driving time, on the effects of CO 2. This second objective builds on the idea highlighted by two recent reviews that CO 2 effects may be more easily detectable when task demands are sufficiently high. Besides learning about the role of CO 2 on driving safety, this project also aimed to provide recommendations to Senseair (partner in the project) on the calibration of future in-vehicle CO 2 sensors.

To address these objectives, a study was designed to be conducted on the VTI simulator in Linköping (Sweden). The simulator was placed inside a 24m 3 tent specially built for this project in order to manipulate and regulate the indoor CO 2 levels. This was done by means of a system that allowed pure CO 2 mixed with fresh air to be injected into the appropriate concentrations. The indoor CO 2 level was automatically regulated through sensors and a closed-control system. Specifically, thirty-two healthy adult participants (41 years old on average) performed two computerized cognitive tasks and drove the simulator under levels of 700 (normal indoor), 1500 and 3000 ppm in counterbalanced order. The driving scenario consisted of a rural road with mild curves and light traffic, thus simulating monotonous driving. Each drive lasted 40 minutes. Every 5 minutes, the drivers were asked to verbally report /her level of sleepiness using the Karolinka Sleepiness Scale (1 very alert - 9 very sleepy). At the beginning or in the middle of each drive, a 20-minute auditory task started and performed simultaneously to the drive. The aim of this task was to increase the cognitive load of the participants in order to see its influence on the effects of CO 2. In addition, their heart and eye activity were monitored by electrodes to analyze their level of alertness during driving.

Briefly, our results indicated that the exposures at 1500 ppm led to the strongest effects on our participants' subjective and physiological state, as well as on their performance. Specifically, objective and subjective sleepiness was higher, and lateral vehicle control was worse (i.e., greater lateral variability and more line crossings), compared to the other conditions. Speed variability was also greater in this condition, although only when the auditory task was included. Under 3000 ppm, the only effect observed was greater objective sleepiness (i.e., longer duration of blinks) compared to the baseline condition. Driving performance, however, was similar and, in some cases, slightly better than the baseline condition. Finally, although significant effects of driving time on most subjective, physiological and performance measures were detected, these were not exacerbated by exposure to elevated levels of CO 2, against our prediction.

This is the first study of its kind to investigate the effects of CO 2 on driver fitness and performance during monotonous conditions. Although the results are inconclusive and need to be replicated, we found evidence that moderate-high concentrations of CO 2 could affect the state and performance of drivers. However, these effects need to be qualified. On the one hand, the size of the observed effects is quite small, so we might expect that CO 2, at the concentrations analyzed, would cause serious performance problems in low complexity traffic conditions. Despite this, recent reviews suggested that the effects of CO 2 on performance can be larger under more demanding tasks. Therefore, the impact of CO 2 on driving performance could be greater in more demanding traffic environments (e.g. in urban driving with a variety of road users around). On the other hand, according to our observations, the effects of CO 2 on performance may not follow a linear dose-response pattern. That is, higher levels of CO 2 may not necessarily be more harmful than lower levels. Although there is no obvious explanation for this phenomenon, other studies have reported similar effects. The activation of ill-defined compensatory mechanisms in response to elevated CO 2 levels (in our case, 3000 ppm) could explain this non-linear pattern. The investment of more mental effort, increments in respiration breath and depth, or the activation of metabolic processes at the cellular level in response to elevated CO2 levels are some hypotheses proposed in the literature that should be considered in future studies.