China
Africa
Explore chapters and articles related to this topic
Machine learning and the economy
Published in Siddhartha Mitra, Robotization and Economic Development, 2023
It must be remembered though that driver-less vehicles are at the extreme end of the spectrum of autonomous vehicles. Usually, five levels of such vehicles are considered with AI able to exercise increasing amounts of autonomy for a higher level of the autonomous vehicle, i.e., perform a greater proportion of the actions used to control the car – for example, steering the vehicle, braking, and accelerating. In all of levels 1–4, drivers are required to intervene under conditions of an emergency, though for a level of 4, under certain ideal conditions the human driver can take a clean break from the task of managing the car and even take a nap. In levels 2–3, the human driver is also assisted by AI through facilities such as collision detection, lane departure warning, and adaptive cruise control, which adjusts the speed of the car to traffic conditions and the speed limit. It is the level 5 vehicles which are fully autonomous – these can be used as fully robotised taxis to deliver food, etc. (Kapoor, 2020).
Security, Privacy, and Trust of Emerging Intelligent Transportation: Cognitive Internet of Vehicles
Published in Mohiuddin Ahmed, Nour Moustafa, Abu Barkat, Paul Haskell-Dowland, Next-Generation Enterprise Security and Governance, 2022
Khondokar Fida Hasan, Antony Overall, Keyvan Ansari, Gowri Ramachandran, Raja Jurdak
More closely, in autonomous vehicles, a broad range of sensors is integrated into the vehicle to collect the data about the surrounding environment enabling them to operate independently. Primarily, autonomous vehicles can be considered as self-controlled robots that operate independently and take a wide range of decisions onboard without human interaction [10]. While traveling on the road, these vehicles can perceive, understand, and interpret traffic scenarios, make intelligent decisions, and act upon them. The ability to operate independently is facilitated mainly by the innovation in robotics and available tools and techniques. Noticeable breakthroughs in robotics have been observed in recent years due to the advancement of Artificial Intelligence (AI). AI enables machines to make independent, intelligent decisions like humans. AI-powered vision and signal processing techniques gather information around the road environment, interpret, and model the information to make necessary decisions. The actions may further change the environment, requiring subsequent decisions. This results in a close-looped system that tightly integrates the physical and digital realms.
Introduction
Published in William Riggs, Disruptive Transport, 2018
With Level 1 automation, the driver remains in control of the vehicle, but the technology can assist the driver by controlling one of the vehicle's functions, either its speed or lane position. Level 2 takes this a step further by allowing the vehicle to control two driving functions at the same time. A vehicle with Level 3 automation can take full control of the vehicle for certain parts of a trip, but drivers must be ready to take back control of the vehicle when the vehicle prompts them. The vehicle takes full control of all major driving functions in Level 4. Level 4 vehicles can even drive themselves for the entire trip, but they are only able to do so under specific conditions. Finally, Level 5 automation refers to fully autonomous vehicles that can operate without an operator in all conditions and without the capability for a human to retake control.
Understanding Passenger Acceptance of Autonomous Vehicles Through the Prism of the Trolley Dilemma
Published in International Journal of Human–Computer Interaction, 2023
Autonomous vehicles refer to artificial intelligence–enabled autonomous, self-driving cars that operate without human control (Biondi et al., 2019; SAE International, 2021). Based on SAE J3016_202104 (SAE International, 2021), artificial intelligence–enabled autonomous vehicles should be classified as highly automated vehicles (level 4) and fully automated vehicles (level 5) that require minimal or no control from the human driver. The development of autonomous vehicles can generate positive social impacts, such as decreasing traffic congestion and transportation cost, enhancing road capacity, offering mobility to people with disabilities and the elderly, and reducing pollution (Anderson et al., 2016; Fagnant & Kockelman, 2015). One of its main benefits is the significant reduction in traffic accidents (Anderson et al., 2016; Rhim et al., 2020; Waldrop, 2015), given that approximately 94% of road accidents are due to human driver errors (Singh, 2018).
Automatic traffic modelling for creating digital twins to facilitate autonomous vehicle development
Published in Connection Science, 2022
Shao-Hua Wang, Chia-Heng Tu, Jyh-Ching Juang
Autonomous driving software is the central control of an autonomous vehicle, and it receives the sensor data and reacts based on the perceived information. The autonomous driving software is considered as a platform that consolidates different algorithms to facilitate the self-driving researches, such as Zhao et al. (2021), Das and Chand (2021), Jiang et al. (2021), and Zhang et al. (2021). In addition to those proprietary solutions, Apollo (Baidu, 2020) and Autoware (Kato et al., 2015) are the open-source alternatives, both of which are highly integrated with the SVL simulator for autonomous vehicle simulations. In this work, we use the Robot Operating System (ROS) based driving software, Autoware, to facilitate the development of traffic modelling; further information is detailed in Section 4; note that ROS is a software framework to create robotic applications. In the real world, Autoware is running within the physical vehicle. On the contrary, in the simulated environment, Autoware can be run in either the co-simulation scheme or the standalone scheme.
Long-term prediction for high-resolution lane-changing data using temporal convolution network
Published in Transportmetrica B: Transport Dynamics, 2022
Yue Zhang, Yajie Zou, Jinjun Tang, Jian Liang
Autonomous vehicles can effectively mitigate traffic congestion, improve traffic safety and reduce vehicle emission pollution. One challenging task of autonomous driving is to model the driving behaviour and then predict the future vehicle movement. As one of the basic driving behaviours, lane change is a complex vehicle movement and can potentially result in serious traffic accidents. Lane change includes vehicle movement in both horizontal and vertical directions, and frequent lane change behaviours can significantly affect traffic flow operation. The national highway traffic safety administration estimates that in 2007, among all traffic accidents reported by the police, lane change and merging traffic accidents accounted for about 0.5% of traffic fatality (Guo, Wotring, and Antin 2010). In addition, studies have shown that dangerous lane changes can cause unstable traffic flow (Yang et al. 2009), and the lane change/merging accidents can also cause significant traffic delay (Chovan et al. 1994). Modelling the lane-changing process is useful for advanced driver assistance system (ADAS) (Nilsson et al. 2017) to implement automatic emergency braking, forward collision warning and lane departure warning, etc. Therefore, accurate prediction models are definitely useful to improve the safety of lane-changing behaviour.