Driver drowsiness is one of the leading causes of traffic accidents. This paper proposes a new method for classifying driver drowsiness using deep convolution neural networks trained by wavelet scalogram images of electrocardiogram (ECG) signals. Three different classes were defined for drowsiness based on video observation of driving tests performed in a simulator for manual and automated modes. The Bayesian optimization method is employed to optimize the hyperparameters of the designed neural networks, such as the learning rate and the number of neurons in every layer. To assess the results of the deep network method, heart rate variability (HRV) data is derived from the ECG signals, some features are extracted from this data, and finally, random forest and k-nearest neighbors (KNN) classifiers are used as two traditional methods to classify the drowsiness levels. Results show that the trained deep network achieves balanced accuracies of about 77% and 79% in the manual and automated modes, respectively. However, the best obtained balanced accuracies using traditional methods are about 62% and 64%. We conclude that designed deep networks working with wavelet scalogram images of ECG signals significantly outperform KNN and random forest classifiers which are trained on HRV-based features.

In this paper, we focus on the rationale for implementing the fare-free public transportation system (FFPTS) in Tallinn, Estonia, that took place on 1 January 2013. Through a series of interviews with relevant informants, we identify the main enablers and the FFPTS in Tallinn faced and the potential of such a system to contribute to the sustainable city development. Our analysis shows that the interlinking between local and national politics determines not only the type of initiatives implemented and the support they receive but also the degree of their success and their stability. We conclude that to be even more effective, it should be extended to the all potential users, not just to local registered residents as it has been recently applied in state-run bus travels in rural municipalities in Estonia. Finally, more restrictive private car policies should be considered to fuel a sustainable mobility transition and increase cities life quality.

User-related assessment of a level 3 automated driving system, providing functions such as lane- and distance-keeping, stop & go driving, lane change and overtaking, was carried out on a motorway in Germany with the aim to assess user-related issues of automated driving, i.e., behavior when driving with automation on motorways, user experiences, reactions and acceptance. Twenty-one persons drove twice along the test route once with the system switched off and once with the system active. Driving data were logged and driving behavior was observed by two observers in the car and the drivers answered questionnaires. The results revealed that the drivers used the system as it was intended to be used and that the system affected driving positively in several ways, resulting in better speed adaptation, less speed variation, better distance keeping, better lane choice, better indicator usage and fewer dangerous lane changes. No differences with regard to subjective workload were shown. The system was perceived as being both useful and satisfactory. However, it was found that the system did not react to other drivers’ intention to make a lane change, especially in situations where they were attempting to merge onto the motorway. Further development is needed to improve the system’s ability to recognize other vehicles’ intention to enter the vehicle’s own lane. Also, the phenomenon of the driver “feeling guilty” because of the system’s reckless behavior by not allowing other drivers to merge onto the motorway or hindering other cars behind due to lengthy overtaking should be investigated further.

The development of automated driving is an ongoing process; nonetheless, certain problems remain unresolved. One of them is the question when the automated vehicle control system should hand over the control to a human driver and whether this ...

Driver drowsiness is one of the leading causes of traffic accidents. This paper proposes a new method for classifying driver drowsiness using deep convolution neural networks trained by wavelet scalogram images of electrocardiogram (ECG) signals. Three different classes were de-fined for drowsiness based on video observation of driving tests performed in a simulator for manual and automated modes. The Bayesian optimization method is employed to optimize the hyperparameters of the designed neural networks, such as the learning rate and the number of neurons in every layer. To assess the results of the deep network method, Heart Rate Variability (HRV) data is derived from the ECG signals, some features are extracted from this data, and finally, random forest and k-nearest neighbors (KNN) classifiers are used as two traditional methods to classify the drowsiness levels. Results show that the trained deep network achieves balanced accuracies of about 77% and 79% in the manual and automated modes, respectively. However, the best obtained balanced accuracies using traditional methods are about 62% and 64%. We conclude that designed deep networks working with wavelet scalogram images of ECG signals significantly outperform KNN and random forest classifiers which are trained on HRV-based features.