China
Africa
Explore chapters and articles related to this topic
Securing Future Autonomous Applications Using Cyber-Physical Systems and the Internet of Things
Published in Amit Kumar Tyagi, Niladhuri Sreenath, Handbook of Research of Internet of Things and Cyber-Physical Systems, 2022
S. Sobana, S. Krishna Prabha, T. Seerangurayar, S. Sudha
In recent years, the concept of self-driving cars has generated significant attention and discussion in various news and reports [13, 14]. Self-driving cars are also termed as the wheeled mobile robot, AV, connected and autonomous vehicle (CAV), driverless car, robocar, or robotic car, [15, 16] and it is a kind of intelligent car [17]. Self-driving cars are able to senses its environment and reaches a destination with a small or no manual input depending upon the information received from automotive sensors, the details include path environment, route information and car control. Without humans’ assistance this car is capable of transporting people or things to a predetermined destination [18]. Various sensors are used to sense the surrounding includes radar, sonar, lidar, GPS, and odometry [15]. Technologies such as automatic control, AI, architecture, computer vision are integrated into the design of self-driving car.
Intelligent Transport Systems and Traffic Management
Published in Rajshree Srivastava, Sandeep Kautish, Rajeev Tiwari, Green Information and Communication Systems for a Sustainable Future, 2020
Pranav Arora, Deepak Kumar Sharma
A self-driving or autonomous car is a vehicle that is capable of sensing its nearby environment and moving on the road in a safe manner with no or negligible human input. Self-driving cars use various kinds of sensors to see or predict their surroundings, such as radar, GPS, sonar, odometry, speedometry, and various other inertial measuring devices. These vehicles are fitted with powerful computers, along with highly sophisticated AI algorithms, consisting of various ML algorithms and applied use of computer vision. The vehicles feed in the information, with the help of various sensors, and processes, in real time, all the scenarios that could possibly occur. The system calculates the speed of the nearby cars, as well as keeping a look-out for pedestrians, traffic signals, and road markings to create and follow a safe route. All control of the vehicle, from acceleration to braking and lane-keeping to turning is handled by the system. Driving safety experts and statisticians predict that once this self-driving car technology has become fully developed, accidents or collisions caused by human error, such as delayed reaction time caused by distraction or aggressive driving styles, will be significantly reduced. In addition to being safe, it will also allow us to travel at a much faster pace, thus reducing the daily commute times and hence making our lives more efficient. It will also provide access for people with disabilities who cannot themselves drive. Figure 3.8 depicts such a system where cars can communicate with one another.
An Overview of Deep Learning in Industry
Published in Jay Liebowitz, Data Analytics and AI, 2020
Quan Le, Luis Miralles-Pechuán, Shridhar Kulkarni, Jing Su, Oisín Boydell
An autonomous car or self-driving car is “a vehicle that is capable of sensing its environment and navigating without any human input” (Hussain, 2016). Deep learning approaches are often used for object recognition as part of an autonomous driving pipeline, and technologies based on these systems dominate the current commercial autonomous driving efforts. Google’s self-driving car unit, for example, started in 2009 and in the next seven years drove over two million miles of test journeys on open roads. This car implements deep learning models extensively for object recognition. Similarly, the Tesla Autopilot system incorporates Tesla-developed deep learning systems based on CNNs for the tasks of recognizing objects through vision, sonar, and radar sensors. There are also examples of smaller startup self-driving car companies such as Drive.ai, which created a deep learning-based software for autonomous vehicles, or Tealdrones.com, a startup that equips drones with onboard deep learning modules for image recognition and navigation.
Interactions between cyclists and automated vehicles: Results of a photo experiment*
Published in Journal of Transportation Safety & Security, 2020
Marjan P. Hagenzieker, Sander van der Kint, Luuk Vissers, Ingrid N. L. G van Schagen, Jonathan de Bruin, Paul van Gent, Jacques J. F. Commandeur
SWOV is interested in how cyclists react when interacting with cars in traffic. In this experiment two sets of 30 photos of different bicycle-car interactions will be shown from the perspective of the cyclist. In random order photos of interactions with manually driven cars and self-driving cars will be shown. By manually driven cars we mean those as they participate in traffic nowadays. With self-driving cars we mean cars that will encounter traffic in the (near) future. These automated cars are vehicles equipped with sensors and computer- and communication systems which carry out the entire driving task. Self-driving cars drive to a predestined location and perform all tasks the driver normally does, like steering, controlling speed, keeping distance, overtaking etc.
Smart Mobility in Smart Cities: Emerging challenges, recent advances and future directions
Published in Journal of Intelligent Transportation Systems, 2023
Soumia Goumiri, Saïd Yahiaoui, Soufiene Djahel
Predicting the evolution of traffic flow in urban areas is more difficult than freeways due to the complexity of the road layout and traffic pattern variations. As opposed to long-term traffic prediction, short-term traffic prediction is more accurate since the used data are updated more frequently at the expense of the required large storage and fast processing. Achieving fast and accurate short-term traffic prediction is a necessity as it enables faster and more efficient response to potential bottlenecks. We foresee that Intel’s Myriad X Vision Processing Unit (VPU)—the first of its class to feature the Neural Compute Engine—providing a dedicated hardware accelerator for deep neural network inferences will enable the development of new highly accurate deep neural network based traffic prediction models. Such models are expected to meet the above speed and accuracy requirements. Responding efficiently to predicted deterioration of traffic conditions may involve alteration of default driving policies or traffic signaling systems through smart and adaptive traffic light control systems, such as the schemes developed by Djahel et al. (2020) and Aleko and Djahel (2020), or virtually inflating road network capacity through granting a selected set of vehicles temporary access to under-utilized reserved lanes Djahel et al. (2018). We believe that the future of traffic management systems is VANET-based. Due to their rapid reactions, self-driving cars would significantly reduce traffic congestion and accidents. However, with VANETs in particular and IoT-based systems in general, privacy, security, and authentication difficulties become crucial challenges.
Why Is Artificial Intelligence Blamed More? Analysis of Faulting Artificial Intelligence for Self-Driving Car Accidents in Experimental Settings
Published in International Journal of Human–Computer Interaction, 2020
Joo-Wha Hong, Yunwen Wang, Paulina Lanz
Self-driving cars are vehicles with automation technologies that do not require human interference for driving (Vellinga, 2017). These vehicles can drive without human input using a decision-making module to understand circumstances based on gathered information from sensors, just like a human driver (T.S. Kim et al., 2012). The development of automation technologies for cars in recent years is significant, and these technologies are expected to be deployed in the near future (Borraz et al., 2018; Lee et al., 2018). Despite the growth of this business, there was a fatality caused by a self-driving car in Phoenix, Arizona, on March 18, 2018; this triggered discourse about the reliability and safety of this artificial intelligence-based technology (Wakabayashi, 2018). In this accident, the blame for the death of the victim could potentially be assigned to the self-driving car (both the car itself and the self-driving technology), the backup driver, environments, or the victim, which is a more diverse set of possible causes compared to those for previous human-driving car accidents. The distinctiveness of this case regarding the attribution of responsibilities is because it is an accident caused by a non-human entity that performs what was formerly considered a human activity. Also, people were not ready to face a death by self-driving cars. It is found that legislators in the US have been slow to regulate self-driving cars, in part due to optimistic views that the adoption of this technology would reduce the frequency of accidents (Browning, 2014). These observations suggest that society in general was not prepared for the recent tragic accident, resulting in diverse reactions and opinions about self-driving cars.