There is still much to learn about how future users will accept different types of electric vehicle (EV) charging.

Key questions include what kinds of charging methods users are comfortable with, such as smart charging or vehicle-to-grid (V2G) solutions. To support better planning of charging infrastructure, electricity providers have commissioned VIRTUAL VEHICLE to explore these topics. Additionally, VIRTUAL VEHICLE is conducting similar investigations as part of a project funded by Horizon Europe.

To gain insights, VIRTUAL VEHICLE carried out a psychologically structured survey and conducted interviews with over 500 participants across Europe, including operators, vehicle owners, and private customers.

The results show that business readiness for EV charging depends on both fixed elements—like the operational requirements of the vehicles—and adaptable factors, such as how the current vehicle schedules are organized. Meanwhile, private consumers showed more hesitation toward vehicle-to-grid or vehicle-to-building options compared to simpler vehicle-to-home solutions.

In cooperation with:

A key challenge in higher-level automated driving (SAE Level 3 and 4) is identifying a reliable way to predict how quickly drivers can take back control of the vehicle, especially when they are engaged in non-driving-related activities (NDRA) or are in a reduced state of alertness. This is important because safety regulations must define which driver postures or actions are too risky and could delay a prompt takeover response.

To study this, researchers measured response time — defined as the time needed to place both hands on the steering wheel after a takeover request. The safety threshold is set at 1.4 seconds, according to EU Commission Regulation 347/2012. A Time-of-Flight sensor was used to detect various driver postures, such as drinking, tapping on a screen, using a smartphone, crossing arms, resting hands on the lap, or having fingers interlocked behind the head.

Machine learning algorithms then analyzed the data to determine the driver’s hand position at the moment of the takeover request. The study found that the distance the hands needed to reach the wheel is a strong predictor of response time — with an accuracy of around 70%, outperforming current state-of-the-art models.

The results also showed that activities involving objects in the hands often led to response times that exceeded the safety threshold. Additionally, drivers who were inactive for just 10 minutes in SAE Level 4 automation already showed slower reactions, likely due to reduced alertness.

In cooperation with:

One of the most critical challenges for drivers using vehicle automation and driver assistance systems is the ability to safely take back control of the vehicle when required. This is known as the “human out-of-the-loop” problem, where the driver may not be fully engaged or ready to react quickly during automated driving.

To address this, a research project supported by Horizon Europe and national funding set out to measure how likely unsafe take-over events are to occur. The project included both controlled driving simulator studies and real-world field tests on a test track, where participants performed take-over maneuvers under the supervision of a safety driver.

The findings revealed that 61% of trials and 64% of actual take-over events showed issues during the hand-over process. Based on this, researchers developed a human error likelihood model that can predict failure rates. The study also highlighted current problems in existing Level 2 (L2) automation implementations, pointing to areas where systems and safety protocols can be improved.

In cooperation with:

One of the key challenges in adopting SAE Level 1 driver assistance systems is that we still know very little about how drivers initially interact with them and how their trust in these systems develops over time. This lack of understanding limits the ability to design better and safer driver assistance solutions.

To address this, a study funded by the FFG investigated how different types of drivers interact with assistance systems and how their trust evolves. In two field studies, involving a total of 137 participants, researchers observed how drivers used these systems in real-world road environments.

The results revealed a number of usage problems and showed that many drivers had inaccurate mental models of how the systems work. The way drivers develop trust appears to be function-specific, meaning they trust some features more than others. In particular, there was a clear tendency for drivers to be more comfortable using Adaptive Cruise Control than systems like Lane Keeping or Lane Change Assistance.

In cooperation with:

Ebinger, N., Neuhuber, N., Moser, J., Trösterer, S., & Stocker, A. (2024). Which partially automated driving function do drivers prefer? Results from two field studies on public highways. Transportation Engineering, 17, 100236. https://doi.org/10.1016/j.treng.2024.100236