Is improved lane keeping during cognitive load caused by increased physical arousal or gaze concentration toward the road center?

Abstract This paper presents a technique for the lane keeping and the longitudinal speed control of an autonomous vehicle with the combination of an MPC and a PID control. The goal of the proposed control method is to minimize the lateral deviation and relative yaw angle with respect to the planned trajectory, while driving the vehicle at the highest acceptable longitudinal speed. The reference profile of the longitudinal speed is computed considering both the lateral and longitudinal dynamic of the vehicle. The vehicle is represented by means of a linear 3-DoF bicycle model. The control algorithm takes the road lane boundaries as the only external input. The proposed strategy is validated in simulation on three distinct driving scenarios.

Download Full-text

Predicting takeover response to silent automated vehicle failures

10.31234/osf.io/bv2pt ◽

2020 ◽

Author(s):

Callum Mole ◽

Jami Pekkanen ◽

William Sheppard ◽

Tyron Louw ◽

Richard Romano ◽

...

Keyword(s):

Cognitive Load ◽

Response Times ◽

Human Capabilities ◽

Bayesian Hierarchical ◽

Performance Metric ◽

Driver Performance ◽

Human Driver ◽

Increased Risk ◽

Lane Keeping ◽

Huge Challenge

It remains a huge challenge to create Automated Vehicles (AVs) that are able to respond safely in all possible circumstances. Silent failures will occur when an AV fails to keep within the safety envelope and does not detect this failure or alert the human driver. To ensure AV safety, it is crucial to have a better understanding of human capabilities responding to silent failures. A highly controlled experiment was conducted to test drivers detecting and steering in response to a range of lane keeping failures of automation, using Time-to-Lane-Crossing (TLC) as the primary performance metric. Bayesian hierarchical modelling was used to construct predictive models that showed drivers responded more slowly (and less consistently) during less critical failures (for each 1 s increase in TLC at failure there was a 0.36 s increase in TLC at takeover). A manipulation that increased cognitive load impaired driver performance further (TLC at takeover decreased by 0.1 s and variability increased by 10\%). Steering response magnitudes scaled according to TLC at takeover, but increased cognitive load dampened steering. Whilst these results demonstrate increased risk caused by additional cognitive load, the magnitude of the effect was fairly small compared to the within and between participant variability. Modelling this variability allowed simulations of hypothetical silent failures to be run based on different road conditions (varied curvature, width and speed) and various delays in response times. This modelling suggests that a high proportion of silent failures would result in unsafe transitions of control from AV to a human driver.

Download Full-text

When moderate mental workload and driving are mutually beneficial

10.31234/osf.io/bg89s ◽

2021 ◽

Author(s):

Fatima Maria Felisberti ◽

Thiago P Fernandes

Keyword(s):

Cognitive Load ◽

Cognitive Performance ◽

Cognitive Processing ◽

Mental Workload ◽

Cognitive Task ◽

Attentional Processes ◽

Subsequent Performance ◽

Short Period ◽

Driving Task ◽

Lane Keeping

Background: High cognitive load during driving is often disruptive and one of the main causes of road accidents. Surprisingly, we know little about the effect (if any) of cognitive load immediately before driving, and even less about the effect of driving (with its own cognitive load) on subsequent performance in cognitive tasks. Method: The effect of cognitive load on a subsequent driving task was examined in Study 1 (n = 31). Participants completed a battery of cognitive tests with low or moderate cognitive demands and their driving performance on a simulator was assessed on two consecutive days (speed, distance from the car ahead, and lane keeping ability). Study 2 (n = 98) examined the effect of a cognitively demanding driving task on the performance of follow up cognitive task, the multi-source interference task (MSIT). In that study, accuracy, and reaction time to MSIT were compared in two conditions: no driving vs post-driving.Results: A moderate level of cognitive load pre-driving led to a modest increase in the distance kept from the car ahead, while a demanding period of driving led to a significant increase in cognitive performance when compared to the control condition (e.g., prior driving).Conclusion: The findings suggest that increases in cognitive processing during periods of demanding mental activity mobilise attentional processes which are likely to remain active for a short period of time benefiting subsequent cognitive performance.

Download Full-text

Lane-Keeping Behavior and Cognitive Load with Use of Lane Departure Warning

10.4271/2017-01-1407 ◽

2017 ◽

Author(s):

Helene G. Moorman ◽

Andrea Niles ◽

Caroline Crump ◽

Audra Krake ◽

Benjamin Lester ◽

...

Keyword(s):

Cognitive Load ◽

Lane Departure Warning ◽

Lane Departure ◽

Lane Keeping

Download Full-text

Comparative Study of End-to-end Deep Learning Methods for Self-driving Car

International Journal of Intelligent Systems and Applications ◽

10.5815/ijisa.2020.05.02 ◽

2020 ◽

Vol 12 (5) ◽

pp. 15-27

Author(s):

Fenjiro Youssef ◽

◽

Benbrahim Houda

Keyword(s):

Deep Learning ◽

Average Distance ◽

Imitation Learning ◽

Learning Loss ◽

The Road ◽

Policy Gradient ◽

Markovian Decision Processes ◽

End To End ◽

Active Research ◽

Lane Keeping

Self-driving car is one of the most amazing applications and most active research of artificial intelligence. It uses end-to-end deep learning models to take orientation and speed decisions, using mainly Convolutional Neural Networks for computer vision, plugged to a fully connected network to output control commands. In this paper, we introduce the Self-driving car domain and the CARLA simulation environment with a focus on the lane-keeping task, then we present the two main end-to-end models, used to solve this problematic, beginning by Deep imitation learning (IL) and specifically the Conditional Imitation Learning (COIL) algorithm, that learns through expert labeled demonstrations, trying to mimic their behaviors, and thereafter, describing Deep Reinforcement Learning (DRL), and precisely DQN and DDPG (respectively Deep Q learning and deep deterministic policy gradient), that uses the concepts of learning by trial and error, while adopting the Markovian decision processes (MDP), to get the best policy for the driver agent. In the last chapter, we compare the two algorithms IL and DRL based on a new approach, with metrics used in deep learning (Loss during training phase) and Self-driving car (the episode's duration before a crash and Average distance from the road center during the testing phase). The results of the training and testing on CARLA simulator reveals that the IL algorithm performs better than DRL algorithm when the agents are already trained on a given circuit, but DRL agents show better adaptability when they are on new roads.

Download Full-text

Impact of Autonomous Vehicles on the Physical Infrastructure: Changes and Challenges

Designs ◽

10.3390/designs5030040 ◽

2021 ◽

Vol 5 (3) ◽

pp. 40

Author(s):

Kareem Othman

Keyword(s):

Autonomous Vehicles ◽

Human Factor ◽

Geometric Design ◽

Pavement Condition ◽

Parking Lots ◽

The Road ◽

Physical Infrastructure ◽

Multiple Issues ◽

Lane Keeping ◽

The Impact

Over the last few years, autonomous vehicles (AVs) have witnessed tremendous worldwide interest. Although AVs have been extensively studied in the literature regarding their benefits, implications, and public acceptance, research on the physical infrastructure requirements for autonomous vehicles is still in the infancy stage. For the road infrastructure, AVs can be very promising; however, AVs might introduce new risks and challenges. This paper investigates the impact of AVs on the physical infrastructure with the objective of revealing the infrastructure changes and challenges in the era of AVs. In AVs, the human factor, which is the major factor that influences the geometric design, will not be a concern anymore so the geometric design requirements can be relaxed. On the other hand, the decrease in the wheel wander, because of the lane-keeping system, and the increase in the lane capacity, because of the elimination of the human factor, will bring an accelerated rutting potential and will quickly deteriorate the pavement condition. Additionally, the existing structural design methods for bridges are not safe to support autonomous truck platoons. For parking lots, AVs have the potential to significantly increase the capacity of parking lots using the blocking strategy. However, the implementation of this parking strategy faces multiple issues such as the inconsistent marking system. Finally, AVs will need new infrastructure facilities such as safe harbor areas.

Download Full-text

Study of Precautionary Assistance System of Lane Keeping Based on Monocular Vision

Applied Mechanics and Materials ◽

10.4028/www.scientific.net/amm.701-702.405 ◽

2014 ◽

Vol 701-702 ◽

pp. 405-412

Author(s):

Li Zong Lin ◽

Chao Huang ◽

Jian Feng Chen ◽

Luo Shan Zhou ◽

Shuai Han

Keyword(s):

High Speed ◽

Fuzzy Controller ◽

Detection Algorithm ◽

Monocular Vision ◽

Assistance System ◽

Memory Space ◽

The Road ◽

Single Output ◽

Edge Features ◽

Lane Keeping

The precautionary assistance system of lane keeping based on monocular vision obtains the road edge features, decides which driving conditions is happened using the fuzzy controller, then gives drivers the corresponding precautionary instructions. In order to improve the processing speed and reduce the needed memory space, the combination of the searching method of connected area and Randomized Hough Transform (RHT) based on two-dimension listing table storage are first proposed in the study, which detects the road edge features. The fuzzy controller is designed with two input and single output, which gives the precautionary instructions in real time. The method proposed in the study is proved that the detection algorithm is simple, and compute velocity is fast, which is fit for the high speed in vehicle driving.

Download Full-text

ROADLANE—The Modular Framework to Support Recognition Algorithms of Road Lane Markings

Applied Sciences ◽

10.3390/app112210783 ◽

2021 ◽

Vol 11 (22) ◽

pp. 10783

Author(s):

Felipe Franco ◽

Max Mauro Dias Santos ◽

Rui Tadashi Yoshino ◽

Leopoldo Rideki Yoshioka ◽

João Francisco Justo

Keyword(s):

Computer Vision ◽

Digital Cameras ◽

Modular Architecture ◽

Driver Assistance Systems ◽

The Road ◽

Lane Marking ◽

Lane Departure ◽

Lane Recognition ◽

Traffic Recognition ◽

Lane Keeping

One of the main actions of the driver is to keep the vehicle in a road lane within its markings, which could be aided with modern driver-assistance systems. Forward digital cameras in vehicles allow deploying computer vision strategies to extract the road recognition characteristics in real-time to support several features, such as lane departure warning, lane-keeping assist, and traffic recognition signals. Therefore, the road lane marking needs to be recognized through computer vision strategies providing the functionalities to decide on the vehicle’s drivability. This investigation presents a modular architecture to support algorithms and strategies for lane recognition, with three principal layers defined as pre-processing, processing, and post-processing. The lane-marking recognition is performed through statistical methods, such as buffering and RANSAC (RANdom SAmple Consensus), which selects only objects of interest to detect and recognize the lane markings. This methodology could be extended and deployed to detect and recognize any other road objects.

Download Full-text

Voice-Activated Dialing or Eating a Cheeseburger: Which is More Distracting during Simulated Driving?

Proceedings of the Human Factors and Ergonomics Society Annual Meeting ◽

10.1177/154193120204600405 ◽

2002 ◽

Vol 46 (4) ◽

pp. 592-596 ◽

Cited By ~ 10

Author(s):

James W. Jenness ◽

Raymond J. Lattanzio ◽

Maura O'Toole ◽

Nancy Taylor

Keyword(s):

Mobile Phone ◽

Driving Performance ◽

Simulated Driving ◽

The Road ◽

Fixed Distance ◽

Minimum Speed ◽

Lane Keeping

We measured simulated driving performance for 26 participants who drove a fixed distance while continuously eating a cheeseburger, operating an automobile CD player, reading directions, or using a voice-activated dialing system to place calls on a mobile phone. Performance was also measured while participants drove without doing other tasks. Participants made the most lane-keeping errors, minimum speed violations, and glances away from the road while reading and while operating the CD player. They made significantly fewer driving errors and glances while voice-dialing the mobile phone or eating, although in both of these conditions they made more driving errors and glances than they did when driving without doing any other activity. We conclude that for simulated driving, placing calls using a voice-activated dialing system is as distracting as eating a cheeseburger, but both of these activities are less distracting than continuously operating a CD player or reading directions.

Download Full-text