Driver Reaction Research Articles

Artificial Intelligence-Empowered Automated Double Emulsion Droplet Library Generation.

Double emulsions with core-shell structures are versatile materials used in applications such as cell culture, drug delivery, and materials synthesis. A droplet library with precisely controlled dimensions and properties would streamline screening and optimization for specific applications. While microfluidic droplet generation offers high precision, it is typically labor-intensive and sensitive to disturbances, requiring continuous operator intervention. To address these limitations, we present an artificial intelligence (AI)-empowered automated double emulsion droplet library generator. This system integrates a convolutional neural network (CNN)-based object detection model, decision-making, and feedback control algorithms to automate droplet generation and collection. The system monitors droplet generation every 171 ms-faster than a Formula 1 driver's reaction time-ensuring rapid response to disturbances and consistent production of single-core double emulsions. It autonomously generates libraries of 25 distinct monodisperse droplets with user-defined properties. This automation reduces labor and waste, enhances precision, and supports rapid and reliable droplet library generation. We anticipate that this platform will accelerate discovery and optimization in biomedical, biological, and materials research.

Optimization of Cellular Automata Model for Moving Bottlenecks in Urban Roads

One of the key reasons why the road capacity of urban roads in China often fails to meet the designed capacity is the mixture of heavy vehicles (slow-moving) and light vehicles (fast-moving). This paper presents a two-lane cellular automaton model suitable for simulating urban road traffic that includes intersections, based on the NaSch model. The model comprehensively takes into account multiple key factors, such as vehicle safety distance, speed differences between adjacent vehicles, the acceleration and deceleration performance of different types of vehicles, and driver reaction time, enabling it to more realistically reproduce the complex characteristics of mixed traffic flows on urban roads. The paper investigates and analyzes the influence of traffic flow density and the proportion of heavy vehicles on the moving bottleneck effect in urban roads from several aspects, including space–time evolution diagrams, traffic flow, average speed, and lane-changing rates. The results indicate that the model established in this paper successfully simulates the characteristics of mixed traffic flows on urban roads, and the simulation results align with actual traffic conditions, achieving the expected simulation effects. Before the traffic volume reaches saturation, the higher the proportion of heavy vehicles on urban roads, the stronger the moving bottleneck effect. This confirms the necessity of conducting research on the phenomenon of moving bottlenecks and the mechanisms of traffic impacts in urban roads, providing a scientific basis for formulating effective traffic dispersion measures and alleviating traffic congestion in the future. This research possesses significant practical meaning and value.

Driving conditions and safety analyses of vehicles moving on highway bridges under seismic excitations

This study introduces an analytical framework and case study to evaluate driving conditions and safety of vehicles on highway bridges subjected to seismic excitation. The driving conditions, namely, driver perceptibility and comfort were examined using overall vibration total value (OVTV) criterion defined in the ISO 2631-1, while driver reaction was simulated using second-order predictable-correction (SOPC) model. A case study was presented using 3D-finite element (FE) model of an existing highway interchange and vehicle-bridge-seismic dynamics (VBSD) model. In this case study, site-specific ground motions at two intensity levels were used to simulate bridge responses. Influences of earthquake intensity, bridge geometry, and road conditions were investigated on five non-articulated vehicles: light vehicle, SUV, small truck, bus, and large truck. Simulations reveal that the driving comfort level for most vehicles has surpassed the uncomfortable and extremely-uncomfortable level defined in the ISO standard under L1 earthquake and L2 earthquake, respectively. Driving conditions worsened on a curved bridges compared to straight bridges. Three driver reaction scenarios were simulated to examine risks and safety indices. The results showed that deceleration action was the most effective response, followed by a combination of deceleration and changing course, while changing course only was the least favorable outcome.

Improving work zone safety: Integrating VR-CARLA co-simulation and eye tracking for behavior analysis of drivers around work zones

Ensuring safety around roadway work zones is difficult which is evident by the fact that despite considerable safety measures, drivers and workers are still exposed to risks. An incomplete understanding of driver behaviors around work zones is one of the contributors to accidents and incidents, prompting the establishment of behavioral rules and a thorough assessment of driver awareness around work zones to enhance safety on roadways. Advancements in digital technologies combined with major developments in traffic simulation tools, including eye-tracking and the Car Learning to Act (CARLA) simulation platform, provides safe and realistic environments to virtually simulate diverse hazardous work zone scenarios to capture driver’s behaviors without actually exposing them to real-world risks. This study presents a multi-module immersive car simulation and interactive driving platform to capture authentic driver reactions and behaviors around unstructured work zones, by leveraging eye tracking and VR technologies combined with the CARLA co-simulator. Two scenarios with three cases of roadway work zone configurations were implemented for conducting user studies, wherein driver awareness while traversing around work zones was assessed through gaze duration and fixation ratios. The findings indicate that drivers prioritize their attention on workers exhibiting risky behaviors over warning signs, with increased emphasis on these individuals despite signage existence. Warning signs enhanced the awareness of normal workers; however, in hazardous conditions, drivers’ attention disproportionately diverted to risky workers, diminishing their focus on normal workers. Different driver awareness patterns observed based on how drivers were notified about work zones suggest that multi-type notification systems, such as dynamic digital displays, wearable devices for workers, and auditory alerts, with human notifiers at the center, be implemented to keep them aware of work zones and workers.

A new car-following model considering the driver's dynamic reaction time and driving visual angle on the slope

A new car-following model considering the driver's dynamic reaction time and driving visual angle on the slope

Auto-Fusing Covariance and Phase Locking Value With Brain-Inspired Spiking Neural Networks for EEG-Based Driver Reaction Time Prediction

Auto-Fusing Covariance and Phase Locking Value With Brain-Inspired Spiking Neural Networks for EEG-Based Driver Reaction Time Prediction

Priority Design in Multi-Target AR-HUD Warning: Evidence from Eye Movement and Behavior of the Novice Driver

Technically, Augmented Reality Head-Up Display (AR-HUD) technology can augment multiple targets in the scene. However, existing literature predominantly focuses on scenarios involving single-target augmentation. The understanding of multi-target augmentation and studies exploring effective presentation methods under such conditions are notably scarce. This study evaluates the efficacy of integrating color-based warning priority design across multi-target scenarios. 45 Participants in different warning modes (Equivalent, Hierarchical, and Baseline) view AR-augmented driving videos and respond to risky targets. Their behavioral performance and eye-tracking data are compared. Findings indicate that the equivalent warning mode, lacking in priority design, adversely affects driver performance, prolongs reaction times, and elevates saccade counts, and gaze entropy. Conversely, the hierarchical warning mode significantly ameliorates driver reaction times and the time to first fixation, while also reducing saccade counts and gaze entropy, demonstrating the efficacy of the warning priority design. The findings provide insight into the design of AR-HUD with multi-target augmentation.

Optimal Adaptive Cruise Control in Mixed Traffic With Communication Latence and Driver Reaction

Optimal Adaptive Cruise Control in Mixed Traffic With Communication Latence and Driver Reaction

Red Light Reaction – A Statistics Project with Real Life Application

We describe a hands-on project in which students collect data on the impact of distracted driving on driver reaction time. Initially they do this in class via a virtual driving applet, using themselves and fellow students as test subjects. Different applet versions simulate driving with and without distraction and measure the time it takes to apply brakes after the red brake lights of the car ahead appear. Students use the collected data to practice a range of statistical techniques, such as assessing data for outliers, creating a hypothesis to be tested, performing an appropriate test, and then translating their results to determine a safe driving distance. In the second part of the project, students work in groups outside of class. Each group recruits a category of test subjects (e.g., athletes, video gamers, STEM majors) of their choosing, collects data, and performs statistical analysis. Finally, students develop hypotheses as to whether different categories of drivers have better or worse reaction times, collect additional relevant data, and perform the appropriate statistical test. The applets, the project guide, the dataset, and supporting videos can be downloaded from https://osf.io/h8yxw/files/osfstorage. For users of the Canvas LMS, we also provide an export cartridge that contains the various project components. Supplementary materials for this article are available online.

Method for Determining the Conditions for Ensuring Stable Rearward Movement of a Semi-trailer Combination with Non-turning Semi-trailer Wheels

The efficiency of organising freight transport and using motor transport in a Smart City depends on a set of its properties, which in the process of operation determine its suitability for use in given operating conditions. Vehicles have different overall dimensions and designs, which determine their maneuverability and directional stability, their ability to make a curve-line movement in the city, as well as to move safely in reverse. Nowadays, tractor-trailer combinations with non-turning semi-trailer wheels are widely used for freight transport. Such a scheme of vehicle construction does not ensure its directional stability when reversing, which can lead to the folding of the tractor and semi-trailer and loss of mobility of the combination. In the absence of additional special systems, the driver controls the rearward movement of the road train using the steering wheel and rear-view mirrors, and the accuracy of delivery to the object depends on the level of the driver's training. The research of driver's reaction time spent on situation assessment, decision making, and realisation has been carried out. It has been revealed that with manual control, there are difficulties connected with training of drivers to estimate the parameters of movement and control of the road train, in particular, with the process of parking and its accurate delivery to the object. With the help of the developed mathematical model of the rearward movement of the road train and the proposed new turn control law, the research of maneuvering of the road train during its delivery to the object has been carried out and the results confirming the possibilities of building an automated system allowing to reduce the influence of the human factor on the parking process, as well as to reduce the time required to perform the maneuver have been obtained. The conducted analysis of system stability with the help of Routh–Hurwitz criterion has determined the conditions of providing stability of the system moving in reverse. The method of controlling the backward movement of the road train and the design of the device realising it has been developed. The results of simulation modeling allowed us to find the necessary values for the control law for the movement of road trains of different lengths, as well as the preferred variants of the initial positions of the road train for the realisation of its precise delivery to the object.

Validating risk behavior in driving simulation using naturalistic driving data

Validating risk behavior in driving simulation using naturalistic driving data

Determination of rates of occurrence for hydroplaning events with naturalistic driving data

Determination of rates of occurrence for hydroplaning events with naturalistic driving data

Attention-based cross-frequency graph convolutional network for driver fatigue estimation.

Fatigue driving significantly contributes to global vehicle accidents and fatalities, making driver fatigue level estimation crucial. Electroencephalography (EEG) is a proven reliable predictor of brain states. With Deep Learning (DL) advancements, brain state estimation algorithms have improved significantly. Nonetheless, EEG's multi-domain nature and the intricate spatial-temporal-frequency correlations among EEG channels present challenges in developing precise DL models. In this work, we introduce an innovative Attention-based Cross-Frequency Graph Convolutional Network (ACF-GCN) for estimating drivers' reaction times using EEG signals from theta, alpha, and beta bands. This method utilizes a multi-head attention mechanism to detect long-range dependencies between EEG channels across frequencies. Concurrently, the transformer's encoder module learns node-level feature maps from the attention-score matrix. Subsequently, the Graph Convolutional Network (GCN) integrates this matrix with feature maps to estimate driver reaction time. Our validation on a publicly available dataset shows that ACF-GCN outperforms several state-of-the-art methods. We also explore the brain dynamics within the cross-frequency attention-score matrix, identifying theta and alpha bands as key influencers in fatigue estimating performance. The ACF-GCN method advances brain state estimation and provides insights into the brain dynamics underlying multi-channel EEG signals.

Enhancing Safety in Autonomous Vehicles: The Impact of Auditory and Visual Warning Signals on Driver Behavior and Situational Awareness

Semi-autonomous vehicles (AVs) enable drivers to engage in non-driving tasks but require them to be ready to take control during critical situations. This “out-of-the-loop” problem demands a quick transition to active information processing, raising safety concerns and anxiety. Multimodal signals in AVs aim to deliver take-over requests and facilitate driver–vehicle cooperation. However, the effectiveness of auditory, visual, or combined signals in improving situational awareness and reaction time for safe maneuvering remains unclear. This study investigates how signal modalities affect drivers’ behavior using virtual reality (VR). We measured drivers’ reaction times from signal onset to take-over response and gaze dwell time for situational awareness across twelve critical events. Furthermore, we assessed self-reported anxiety and trust levels using the Autonomous Vehicle Acceptance Model questionnaire. The results showed that visual signals significantly reduced reaction times, whereas auditory signals did not. Additionally, any warning signal, together with seeing driving hazards, increased successful maneuvering. The analysis of gaze dwell time on driving hazards revealed that audio and visual signals improved situational awareness. Lastly, warning signals reduced anxiety and increased trust. These results highlight the distinct effectiveness of signal modalities in improving driver reaction times, situational awareness, and perceived safety, mitigating the “out-of-the-loop” problem and fostering human–vehicle cooperation.

A Sequence-to-Sequence Car-Following Model for Addressing Driver Reaction Delay and Cumulative Error in Multi-Step Prediction

A Sequence-to-Sequence Car-Following Model for Addressing Driver Reaction Delay and Cumulative Error in Multi-Step Prediction

Large Scale Evaluation of Normalized Hard-Braking Events Derived from Connected Vehicle Trajectory Data at Signalized Intersections, Roundabouts, and All-Way Stops

Intersection safety has been traditionally evaluated using three to five years of crash data. Recent literature suggests that connected vehicle (CV)-derived hard braking (HB) events can provide a surrogate for crashes with only a few weeks or months of data collection. This study used CV trajectories to derive HB events. Then, the HB events were normalized as the ratio of HB events to sampled CV trajectories. The normalized HB ratios were evaluated and compared at 435 signalized intersections, roundabouts, and all-way stops in Indiana. The analysis showed that signalized intersections and roundabouts had the highest counts of HB events, and all-way stops had the highest HB ratios. Through movements at signalized intersections showed the lowest HB ratios, whereas left turns at all-way stops had the highest ratios. A density analysis of the geospatial occurrence of HB events concluded that they tend to occur closest to the intersection center at all-way stops, but are more evenly distributed at signalized intersections. Additionally, a speed analysis indicated that HB events at signalized intersection through movements tend to occur at higher speeds, roughly between 26 and 36 MPH, perhaps due to the driver reaction during the onset of yellow. The findings presented in this study provide transportation agencies with insights on the occurrence of normalized HB ratios at three different intersection types. The data provided in this paper provide a framework for agencies to use HB ratios to screen different types of intersections for further evaluation.

Bus drivers’ anger and anger expression in driver–passenger conflicts

Bus driver anger due to passenger misbehavior may lead to serious bus accidents, yet there are no mature tools available to capture bus driver reactions during driver–passenger conflicts. The bus driver anger scale (BDAS) and the bus driver anger expression inventory (BDAX) were developed to measure the anger levels and expressions of drivers in such conflicts. A bus driver anger model was developed based on 400 questionnaires in Changsha, China. The findings indicate that drivers are most likely to be angered by passenger violations and quarrels among passengers. Drivers irritated by passenger irregularities tend to employ personal aggressive expression or adaptive/constructive expression. Disputes among passengers may lead drivers to resort to unreasonable methods of venting their emotions. Moreover, passengers’ rude behaviors can trigger bus drivers’ aggressive personal expressions. Therefore, it is necessary to establish passenger regulations and encourage drivers to express their anger reasonably in driver–passenger conflicts.

Modeling Pedestrian Near-Crash Events at a Rectangular Rapid Flashing Beacon-Controlled Intersection Using Video Analytics and Long Short-Term Memory Neural Network

Pedestrian safety is a long-standing issue in urban areas, where pedestrian near-crash events are more frequent than in suburban or rural areas. To address the pedestrian safety problem, a proactive approach was explored to assess and predict the severity of these events, which are valuable indicators of potential crashes. Object detection and tracking techniques were used to establish the temporal relationship of pedestrian near-crash events involving vehicles at an intersection controlled with rectangular rapid flashing beacons. A long short-term memory (LSTM) neural network model is proposed to warn a driver 2 s before the vehicle reaches the conflict zone. However, this scenario can be considered optimistic, as the 2 s interval represents an ideal driver’s reaction time, which is more likely to happen in a connected and automated vehicle environment where vehicles receive real-time information about their surroundings and perform some basic tasks such as braking without waiting for the driver reaction. The results demonstrate the effectiveness of the proposed LSTM neural network model, with an area under the curve value of 78.5% on the training data and an overall recall of 71.1% on the test data. The practical significance of this model is its potential to provide timely information about near-crash events, thereby enhancing pedestrian safety at critical points such as intersections.

ENHANCING ROAD SAFETY FOR VULNERABLE ROAD USERS – THE ROLE OF AUTONOMOUS EMERGENCY BRAKING SYSTEMS

Vulnerable road users are the largest group of road accident victims in the world. At the same time, it should be noted that most road accidents are caused by the motor vehicle driver. An opportunity to increase the level of road safety is emergency braking systems installed in vehicles. Their task is to detect the risk of a collision with another object, issue a warning to the driver and, in the absence of reaction, perform emergency braking. The publication presents the results of research conducted by the Motor Transport Institute as part of the PEDICRASH project. As part of the work, a series of tests of the autonomous emergency braking system were carried out to check unusual but probable cases of events in which it should work. The tests were carried out in a closed area using a few selected models of popular passenger car brands. An additional data acquisition system and dummies imitating vulnerable road users: pedestrians and cyclists were used. The constructions represent the silhouette of an adult person and meet the requirements of the Euro NCAP AEB Protocol. The time of the collision at the time of issuing a warning to the driver was analyzed. Assuming that the vehicle was moving uniformly, the distance from the obstacle was calculated at the time of issuing the warning and after the driver's reaction time (0.7 s and 1 s were assumed) from this signal. The deceleration necessary to brake the vehicle was calculated and it was determined whether the driver would have a chance to brake the vehicle before a pedestrian or a cyclist in a given situation. For cases in which the driver would not be able to brake in time, the speed of the vehicle with which it would hit the vulnerable road users was calculated. Possible injuries and accident costs were then estimated.

Evaluation of Driver Reaction to Disengagement of Advanced Driver Assistance System with Different Warning Systems While Driving Under Various Distractions

Advanced driver assistance systems (ADAS) as a growing technology are expected to improve drivers’ performance by carrying out some of the drivers’ tasks and utilizing driver monitoring and warning systems to maintain their awareness. In this study, drivers’ reactions to the disengagement of ADAS and the effectiveness of steering wheel and face tracking warning systems were evaluated using driving simulation. The study was designed as a mixed design experiment to compare the effects of different types of driver monitoring and warning systems on driver response time, when drivers were driving under audio-, visual-, or no distraction. Data from 60 drivers were collected from the driving simulator experiments. In all experimental scenarios, the participant started driving on a long, straight highway segment and activated an ADAS, which was disengaged at the 11th min into the experiment without the driver’s knowledge. The driver’s response to the disengagement was collected and analyzed. A two-way mixed analysis of variance showed that the warning systems and distractions together affected drivers’ response times significantly. Moreover, post hoc test results showed that under the no distraction condition, the mean response time was lower when the face tracker alert was in use compared with no alert, and the response time was significantly lower when drivers were under audio distraction compared with visual or no distraction.