Elsevier

Big Data Research

Volume 17, September 2019, Pages 35-44
Big Data Research

Systematic Review of the Literature on Big Data in the Transportation Domain: Concepts and Applications

https://doi.org/10.1016/j.bdr.2019.03.001Get rights and content

Highlights

  • Big Data and analytics can help garner insights into transportation data.

  • Big Data and analytics can guide the creation of safer transportation systems.

  • Connected vehicles offer great potential to advance transportation systems.

Abstract

Research in Big Data and analytics offers tremendous opportunities to utilize evidence in making decisions in many application domains. To what extent can the paradigms of Big Data and analytics be used in the domain of transport? This article reports on an outcome of a systematic review of published articles in the last five years that discuss Big Data concepts and applications in the transportation domain. The goal is to explore and understand the current research, opportunities, and challenges relating to the utilization of Big Data and analytics in transportation. The review shows the potential of Big Data and analytics to garner insights and improve transportation systems through the analysis of various forms of data obtained from traffic monitoring systems, connected vehicles, crowdsourcing, and social media. We discuss some platforms and software architecture for the transport domain, along with a wide array of storage, processing, and analytical techniques, and describe challenges associated with the implementation of Big Data and analytics. This review contributes broadly to the various ways in which cities can utilize Big Data in transportation to guide the creation of sustainable and safer traffic systems. Since research in Big Data and transportation is, by and large, at infancy, this article does not prescribe recommendations to the various challenges identified, which also constitutes the limitation of the article.

Introduction

Research in Big Data and analytics offers opportunities to apply the evidence-based approach to decision-making in many domains. In the transportation domain, Big Data has the potential to improve the safety and sustainability of transportation systems. Many cities have installed monitoring equipment, such as cameras, roadside sensors, and wireless sensor networks, to observe traffic conditions and promote traffic safety. This equipment collects a massive amount of traffic data and enables transportation departments to gain a better understanding of traffic flow in the respective areas. The availability of traffic data allows both historical and streaming data analysis, which can reveal meaningful traffic patterns, identify congestion, and assist in understanding the causes of collisions or near misses. The various forms of analytics and approaches employed in Big Data, such as machine learning, can be used to sift through the vast amount of traffic data to extract useful knowledge and enable the transportation authority to take preventive actions and make appropriate decisions.

Traffic data contains hidden values that can improve and support safe and sustainable transportation systems. For instance, by using roadside sensors, vehicle speed data can be collected and analyzed to identify traffic congestion. When traffic congestion is detected, travel alerts can be provided to drivers to help them find alternate routes and hence reduce the congestion [1]. Analysis of vehicle wait times at traffic lights can produce insightful information and lead to better ways to optimize traffic light policies and improve traffic flow [2]. Analysis of video data can detect and classify objects (e.g., vehicles or pedestrians), identify their trajectories, and recognize significant traffic events, such as veering, abrupt braking, and near misses [3]. Such analysis can help decision makers take the necessary actions to improve road safety, prevent collisions, and save lives.

Traffic data, arguably, fits the characteristics of Big Data, often characterized along the following dimensions: volume, variety, velocity, veracity, and value [4], [5]. First, various equipment installed on roads to monitor traffic generates a vast amount of data. The volume of traffic data will grow more significantly when connected vehicles communicate and exchange information with other devices within themselves, with the road infrastructure, or with other nearby vehicles. Connected vehicles could generate approximately 30 gigabytes of data per day [6]. At this rate, the traffic data would exceed a terabyte of data over approximately one month. Second, traffic data comes in a variety of structured and unstructured data, such as JPG, JSON, XML, GPS, PDF, image, video, and social media posts [7]. Third, the velocity of traffic data is substantial as various sources produce new data continually. Fourth, data veracity refers to uncertainties that are inherent in traffic data, such as inaccurate or incomplete data [8]. Finally, traffic data contains invaluable information but at a low density. For example, video data may reveal the cause of a collision at an intersection. However, since collisions do not occur all the time, most of the data capture only normal vehicle movement in the area.

Some applications in the transportation domain, such as autonomous vehicles, require real-time data processing and reliable communication networks. Considering the volume of traffic data, Big Data and analytics can benefit from edge computing, which allows data processing and computation to happen near the data sources (e.g., cameras, sensors, mobile devices, vehicles), thereby reducing the bandwidth consumption and network latency between the end-users and the cloud computing platforms that store, manage, and analyze data [9], [10], [11], [12]. Edge computing has the potential to offer an efficient way to tackle issues relating to the exponential growth of data, limited communication bandwidth, and high computational resources in the cloud.

The primary goal of this review is to gain knowledge of the current research and applications of Big Data in transportation. The review is intended to give researchers and transportation departments insights into Big Data and analytics to guide and support the development of better transportation systems. Additionally, this review can assist cities that have adopted Vision Zero [13] in working towards eliminating traffic fatalities and major injuries (i.e., collision injuries that result in admission to hospital).

We organize the rest of this article as follows. Section 2 describes the systematic review protocol including research questions, search terms used, inclusion/exclusion criteria, databases examined, and articles included in this review. Section 3 reviews architectures of Big Data and intelligent transportation systems. Section 4 presents opportunities relating to the utilization of Big Data and analytics to support sustainable transportation systems, whereas Section 5 describes the associated challenges. Section 6 discusses our point of view related to the research questions. Finally, Section 7 summarizes the main points of our literature review.

Section snippets

Systematic review protocol

A systematic review protocol helps researchers to develop a high-level overview of knowledge on a particular research area [14]. It provides a methodical process of identifying, screening, and synthesizing a body of published work in pursuit of holistic evidence relevant to particular research questions. We employed a systematic review to provide a useful overview of the current body of research on Big Data and analytics in the transport domain. This section describes the protocol we developed

Big data systems and their utilization in transportation

This section details how Big Data architectures and systems attempt to address some of the challenges and realize the potential of Big Data in transportation. The research questions addressed in this section are: (1) What is the current state of scientific research on the application of Big Data and analytics in DOTs? (2) What kinds of Big Data systems are being used by DOTs? (3) How are DOTs using insights derived from Big Data and analytics? (4) How can DOTs use Big Data and analytics to

Opportunities

Connected vehicles represent a significant opportunity for Big Data in transportation in the areas of safety and sustainability. A connected vehicle uses GPS, communication technologies, and sensors to communicate with devices within the vehicle and with other vehicles, as well as with transportation networks [26]. Data collected from these sources may be utilized for the internal operation of systems or shared with others for collective benefits.

A connected car can offer personalized services

Challenges in using big data in transportation

Given the nature of transportation Big Data, there are several challenges identified in the literature. These challenges can be summarized as data collection, quality, storage, and security issues (see Fig. 4).

Discussion

The primary purpose of this literature review is to explore the research and applications of Big Data in the transportation domain. We aim to examine the current practices on the use of Big Data and analytics to improve the efficiency and sustainability of transportation systems, rather than to identify and prescribe specific solutions to the challenges of Big Data in transportation. During the analysis of the literature, it became apparent that Big Data projects in the transportation domain

Conclusion

Public transport systems are essential determinants of quality of life. Many countries face challenges in attempts to improve the quality of their public transport systems. Developing efficient and sustainable policies is key to durable transport systems. A report by the International Transport Forum [46] indicated that the growth and availability of vast amount of data in the transportation domain could lead to new policy-relevant insights and operational improvements of traffic. In

Acknowledgements

The project is part of institutional collaboration between MacEwan University, the City of Edmonton, and the University of Otago. We thank Elizabeth Cayen for proofreading an earlier version of this article.

References (46)

  • G. Kemp et al.

    Aggregating and managing big realtime data in the cloud – application to intelligent transport for smart cities

  • A. Artikis et al.

    Self-adaptive event recognition for intelligent transport management

  • W. Shi et al.

    Edge computing: vision and challenges

    IEEE Int. Things J.

    (2016)
  • W. Shi et al.

    The promise of edge computing

    Computer

    (2016)
  • Y. Mao et al.

    A survey on mobile edge computing: the communication perspective

    IEEE Commun. Surv. Tutor.

    (2017)
  • W. Yu et al.

    A survey on the edge computing for the Internet of things

    IEEE Access

    (2018)
  • The vision zero network

  • J. Chandler et al.

    Cochrane methods – twenty years experience in developing systematic review methods

    Syst. Rev.

    (2013)
  • M. Volk et al.

    Providing clarity on big data technologies: a structured literature review

  • Gartner Inc.'s 2012 hype cycle for emerging technologies [online]. (Accessed 24 September...
  • Y. Lin et al.

    Intelligent transportation system (its): concept, challenge and opportunity

  • C. Costa et al.

    Towards real-time road traffic analytics using Telco Big Data

  • P. Ta-Shma et al.

    An ingestion and analytics architecture for IoT applied to smart city use cases

    IEEE Int. Things J.

    (2018)
  • Cited by (83)

    View all citing articles on Scopus

    This article belongs to Special Issue: Smart City Big Data.

    View full text