Developments in connected and automated vehicles

Introduction

Connected and Automated Vehicle (CAV) technologies are a natural extension of Intelligent Transportation System (ITS) initiatives that started in the early 1980s as a result of major advancements in the communication and computation areas. In the United States, the Intermodal Surface Transportation Efficiency Act of 1991 (ISTEA) generated the Automated Highway System (AHS) program (FHWA, 1994) and a concept demonstration in 1997 that was held in San Diego, California (FHWA, 1997). However, the major push forward in the area of connected vehicles happened with the proliferation of GPS-enabled in-vehicle navigation devices combined with mobile communication technologies in the 2000s. Around the same time, USDOT has recognized the potential of connected vehicles being able to communicate with each other as well as with the infrastructure and created a number of programs to expedite the deployment of such technologies (ITS JPO, 2017).

Although the primary focus of USDOT's connected vehicles program has been in safety applications aimed at reducing crashes on roadways, such technologies also have the clear potential of reducing congestion and improving mobility. The combination of all these potential benefits promised by emerging connected vehicle technologies attracted the attention of private sector, especially car makers and technology companies (The Economist, 2009). Connected vehicle technologies also fueled the emergence of a new generation of technology companies that are based on the concept of “mobility as a service.” This concept is bringing completely new business models to the transportation market with the understanding of the change in the trends in the mobility industry (Viereckl, Ahlemann, & Koster, 2016).

More recently, advances in machine learning, e.g., deep learning, and computation made it possible to develop real-time image recognition systems that are crucial to the deployment of automated vehicles (NHTSA, 2017). Several technology giants have been investing in this area with the ultimate goal of developing the automated vehicle of the future. Government agencies and vehicle manufacturers are also in this race for the future of transportation, and they are closely working with these nontraditional stakeholders by participating in several large-scale projects in various parts of the United States, Europe, and Asia.

CAVs will likely revolutionize the way we travel, through the use of advanced communication and computing technologies. Clearly, wireless technologies are in the forefront of this revolution that will transform our transportation system, but other technological advances such as sensors are also going to play a major role, becoming an integral part of vehicles, infrastructure, and people. Moreover, advances in social media and individualized apps are some of the most important aspects in the creation of a new mobility eco-system. In this new eco-system, individuals could directly improve their travel behavior and means by sending customized signals that are designed to improve overall system operations. This is a new concept that transcends vehicles and infrastructure and converges to a relatively new concept of “connected travelers.” Various stakeholders from government, industry, and academia have invested resources to work on topics related to CAVs. A large number of automobile manufacturers and other transportation service providers are leading the way in developing and introducing new concepts and products. The evolution of a highly connected system is also expected to lead first to a partial vehicle automation system and then full vehicle automation. Thus, there is a need to understand the impacts of CAV technologies since transportation touches almost every aspect of modern societies.

Developments in CAV technologies

This special issue is an attempt to bring the latest theoretical ideas and empirical results in CAV together with the goal of contributing to the ever-increasing race toward the transformation of our brick and mortar transportation system to a 21st century intelligent system. The special issue consists of six peer-reviewed articles, selected through a rigorous process, which examines CAV systems from a variety of perspectives.

The article by Zheng, Ran, and Huang (2017) develops a bidirectional cellular automation model to describe car following and driving behavior under the vehicle-to-vehicle (V2V) communication environment. The research is motivated by the fact that under V2V, vehicles can receive information from not only preceding vehicles but also following vehicles, creating an active bidirectional looking context. This is very different from traditional car following models that assume that vehicles can only respond to behavior of preceding vehicles. The authors evaluate their new model for safety performances, and conclude that considering bidirectional information under V2V can help improve safety. More importantly, traffic flow state and drivers' risk-taking preference have an important impact on the safety performance. The proposed methods and findings from this research are expected to shed useful lights on how to improve and evaluate the safety outcomes of V2V and connected vehicles.

Cao et al. (2017) propose a mandatory lane-changing (MLC) model for automated vehicles in urban arterials. MLC is crucial for lane-specific routing planning of automated vehicles. The authors develop a new, optimization-based method to determine the best positions at which an instruction is given to the vehicle for MLC. The model considers the distribution of time spent waiting for safety headway and lane-specific travel times, and applies the shockwave theory and queuing theory to derive the distributions and travel times. The proposed method is validated in simulation, and further tested under various scenarios. The evaluation results show that the lane-changing instruction should be given earlier in the case of denser traffic or a higher travel speed, and vehicles can gain considerable time, especially in the case of lower travel speed in the target lane, by following instructions provided by the proposed model. The proposed MLC model can facilitate the development of a full-scale routing planning framework for automated vehicles.

The article by Naranjo, Jiménez, Anaya, Talavera, and Gómez (2017) examines a detection and warning system aiming to reduce motorcycle crashes using vehicle-to-everything (V2X) communications. The system takes into account characteristics specific to motorcycles such as size, mobility, and movement type. An application for alerting driver of the presence of a motorcycle nearby the vehicle has been developed, tested, and implemented in a mobile device. The warning system builds upon the fact that each vehicle uses a vehicular short-range communications (DSRC) module, which allows a Vehicular Ad-Hoc Network (VANET) to be established, specifically designed to enable V2V communications between vehicles at high speed in order to serve all applications for assistance and driving safety. Testing priority of the different messages in accordance with levels of danger related to the warning and user acceptability are the next steps to further develop this warning system.

In the article by Olia, Abdelgawad, Abdulhai, and Razavi (2017), authors study the optimal number and locations of roadside equipment (RSE) units for freeway travel time estimation under the connected vehicle environment. An optimization problem is formulated to simultaneously minimizing the travel time error rate and the number and locations of RSE units. It is solved by the nondominated sorting genetic algorithm to produce a Pareto front of optimal solutions representing the best possible compromises of different objectives. Simulation results show that the accuracy of the estimated travel time is a function of the RSE placement (both the number and locations). In particular, the error rate is higher when only a few RSE units are placed and the error decreases as more RSEs are deployed. After deploying a certain number of RSEs, adding more RSEs will not significantly improve the travel time estimation results. Furthermore, closely placed RSEs are required for more congested freeway segments. The results are consistent with freeway travel time estimation using traditional detection techniques such as loop detectors (Bartin, Ozbay, & Iyigun, 2007; Ban, Herring, Margulici, & Bayen, 2009; Danczyk & Liu, 2011).

The article by Kim and Peeta (2017) presents a vehicle-to-vehicle (V2V) communication–based advanced traveler information system (ATIS) in the spirit of some of the early papers published dealing with similar V2V and V2I ATIS (Mudigonda, Fukuyama, & Ozbay, 2013). The paper models the dynamic flow propagation of multiple units of information using a multilayer framework that captures the dynamics of the three interacting layers, namely, physical traffic flow, intervehicle communication, and information flow. Similar to the earlier approaches in (Ozbay & Indirakanti, 2010) and (Goel, Imielinski, & Ozbay, 2004) proposed to model propagation of information in highly dynamic traffic systems proposed, the framework presented in this paper describes how the dynamic flow propagation of multiple units of information can be mapped from the traffic flow dynamics and the intervehicle communication constraints. By using synthetic experiments, authors generate insights into the flow propagation of multiple units of information. Some of the most notable conclusions of this paper is the potential of the proposed framework to be used to develop fully decentralized V2V communication-based routing, relying on vehicle-level information and hybrid systems combining centralized and decentralized information generation and dissemination strategies.

The final article in this special issue, co-authored by Milakis, van Arem, and van Wee (2017), discusses the potential policy and societal effects of automated driving according to findings presented in literature and identified future research areas. A ripple effect model is employed to conceptualize the potential sequential impacts of automated driving on several aspects of mobility and society. The first-order implications focus on traffic, travel cost, and travel choices of automated driving; the second-order implications explore vehicle ownership and sharing, location choices and land use, and transport infrastructure; and the third-order implications assess energy consumption, air pollution, safety, social equity, economy, and public health. Based on the literature reviews, authors conclude that vehicle automation would have beneficial first-order impacts—travel time savings, increase road capacity, and increase travel demand. The magnitude of these benefits will likely increase with the level of automation and cooperation and with the penetration rate of these systems. One of research needs identified is to examine the impact of vehicle automation on individual components of travel effort that include comfort, travel time reliability, and utilization of travel time while on the move.

Summary

The articles published in this special issue provide novel methods related to various aspects of CAVs, ranging from individual vehicle control to traveler information generation, and acquisition to the role of VANETs in the CAV environment. Clearly, the future of CAVs is full of new and exciting theoretical and practical developments. Especially in the context of “Smart and Connected Cities” emerging as one of the most promising developments of the 21st century, CAVs have a major role to play. Many transportation agencies such as USDOT are also working very closely with various cities and other stakeholders to create real-world test beds to ensure the timely deployment of CAV technologies. In the near future, new CAV deployment sites, in addition to the ones already funded by USDOT in New York City, Tampa, and Wyoming (https://www.its.dot.gov/pilots/), are anticipated to be active throughout the world. It is our hope that the early CAV-related research results presented in papers such as the ones published in this Special Issue will provide the theoretical basis for successful CAV deployment projects by enabling cities and transportation agencies to obtain maximum benefit from their respective deployments.

Acknowledgments

The co-editors of this special issue would like to thank anonymous referees for their hard work in meticulously reviewing the papers submitted for possible publication. They also thank the Editor-in-Chief, Prof. Asad Khattak, for his help during the initiation and review process of this special issue.