Data is essential to the planning, delivery and management of transport services and infrastructure - data covering home and work locations, leisure destinations and demand for travel between all these and others as well. Data is also necessary for ensuring the safe operation of traffic and understanding and addressing crash patterns and trends, as well as responding to such incidents in real-time. Increasingly, vehicle and map data will become essential for supporting higher and higher levels of automated driving. Transport agencies have been collecting data from an array of sources. The data generally falls into four known applications: traffic volumes and flows (counts), network travel times and traffic speeds (historic and in real-time), incident detection and trip origin-destination matrices. In order to produce these information applications, governments have at their disposal a number of instruments and mechanisms for collecting transport data. However, collection of this type of data has usually been time-consuming and not immune to the characteristic trade-offs between collection costs, coverage and accuracy. Construction of origin-destination matrices, which are an essential input for planning the development of the transport network, or even modifications to it, are generally based on traditional household travel surveys. These surveys are complex, requiring the calculation of large samples and a logistical setup for its distribution and collection, which can be costly and time consuming. Examples like these abound, where the collection mechanisms traditionally known and utilised by the governments are not always keeping up the new technological innovations, nor are adapted to capture the rapid evolution of trends and behaviours within cities. Much of this data has a geospatial component , as well as a temporal component, that allows for a more detailed understanding of where people are, where they are travelling, in what conditions and in some cases how and for what purpose , all this throughout different times of the day. Much of this data is being gathered in new ways, from a broadening array of sensing platforms and in a wide range of formats, with several recognised advantages over traditional methods, such as: scale (coverage of entire transport networks - e.g. road and public transport), data collection latency and frequency (24 hours a day for 365 days a year, and in many cases real-time collection). This data is collected, stored and exploited by a diverse set of actors that extends well beyond the field of transport, and, especially to the private sector. All of these developments enable the delivery of location-based services (LBS). As the private sector continuously collects millions and millions of data points as part of their business models, or as a by-product of the location-based services they provide, the share of mobility-relevant data collected by the private, as opposed to the public sector, is growing and starting to create a considerable gap. But if these gaps are to be closed (and this is debatable), new relationship models and partnerships will be needed. Prior work undertaken by the at the ITF in the context of its Corporate Partnership Board (CPB - see Box 1) discussed these changes and noted that one of the key challenges facing authorities was how to manage the delivery of public policy with increasingly privately sourced and owned data concerning the location and movement of individuals. The benefits of closing this gap are potentially quite large, given efficiencies which could be leveraged through new services with this data. The limited number of applications listed before is mostly constrained by the type of data collected and the mechanisms used to collect it. New applications abound, e.g. using indoor location fixes to determine pedestrian flow patterns or waiting times at stations, or using vehicle occupants’ mobile device accelerometer data to help identify pothole locations through vibrations patterns. In the near future it is foreseeable that many other applications will emerge as new business models are built around them, and if these benefits are to be realised, public and private sector incentives towards sharing of this data should be better aligned. This type of location data can be very useful for managing transport networks and planning for new capacity. It can complement existing data collection and, in some cases, even replace several of the traditional data collection methods at a fraction of the costs. But location data is highly personal and difficult to robustly anonymise, and there are real questions as how to balance data privacy and the benefits that can be derived from innovative uses for this data for managing and helping plan for transport activities. As a next step of the data-related work of the CPB, a workshop entitled “21st Century Public Interest Data Sharing” was held in Paris in November 2015, in order to analyse some key issues in more detail and to involve a large variety of experts and stakeholders in this domain. This workshop addressed the management of location data privacy. It explored whether there is a need for new models framing access to, and use of, mobility-relevant location data and if so, what they might be. It also looked at aspects of public policy as they relate to access and control of transport-relevant data by addressing the following questions: * “Privacy-by-Design” principles are unevenly or not at all incorporated into location data collection. Should this change and how might this impact the usability of location data for traffic operations, planning and safety applications? * What strategies exist to durably protect sensitive personal location data? Should more personal control of individuals’ location and mobility-related data be offered or mandated? If so, how? * What are the broader public policy implications of a switch to more and more private control and ownership of transport-relevant data? * Is there a need to move beyond the current supplier-client relationship governing public authority access to most privately collected location and mobility data? If so, what form might this relationship take? * Is there a benefit from minimum public interest data sets for transport operations, planning and safety applications? If public interest data sets were to be operationalised, how might they be specified and what are the technical challenges in establishing them? * Currently, there is increasing pressure for public data sets to be open for public use. Can open location and mobility data be reconciled with increased data and privacy protection? Background There has been an explosion of data resources in the transport sector in the past decade, particularly data resources relating to location and activities of individuals. These datasets are already being used by innovative commercial actors, transport authorities and other government agencies for a variety of purposes. Whilst many understand the vast potential of using these data resources, many are also searching for a more precise understanding how this potential can be achieved. In particular, how to overcome the many “small data” and “big data” challenges inherent in conjoining multiple disparate data sets in such a way as to extract trustable and useable information. Major challenges to this exist in the fields of privacy, trust, and security around this data from the point of view of individuals, organisations, and governments. The workshop examined two broad discussion themes: * What are unique issues arising in the context of location and mobility-specific data that require special attention regarding privacy, trust, and security? * Much of this data is produced by commercial actors and is central to many, but not all, of their core business strategies. This leads to a situation where the most timely, accurate and helpful data to carry out public policy is no longer held by the public sector mandated to carry out this mission. What then is the new data access and sharing model that should emerge to allow public and private interaction in sourcing, accessing or otherwise co-creating data necessary to manage transport activities and plan for transport networks? Traditional relationships amongst industry, technology, government and citizens are changing. Consequently, roles separating production, service provision, regulation, labour, and consumption are blurring as well. Technology is oftentimes driving these changes, particularly by enabling the emergence of new, often disruptive, in many sectors, but also by allowing more traditional services to innovate and further develop. In this broad context, the importance of citizen-led change is growing. Some governments increasingly feel they no longer have the right tools or sufficient information to accompany these changes and to deliver on public policy objectives. Large quantities of data are being generated and are increasingly available to governments from the commercial sector crowding out more traditional data collection methods employed by transport authorities. With this data comes new and augmented challenges. These include ensuring sufficient in-house technical capacity to use these disparate and oftentimes unstructured data sets and ensuring adequate and inviolable privacy protection for both individuals and commercially sensitive information. Furthermore, the very rapid pace of innovation in the private sector often outstrips regulators’ attempts to keep up with changes in technological developments and new services. Some directly or indirectly transport-related start-ups begin quite small and with little budget, then rapidly acquire external funding through venture capital and other sources and grow into large companies. Once established, these companies have a significant and lasting impact on transport behaviour and behaviours that impact transport demand. Transport data emerging from and around these ventures, particularly geolocalised data, must be seen within a broader context, especially in view of the wider digital enablement of society as a whole. Data fuels both innovation and disruption. A key challenge to be addressed revolves around data ownership and use. Data relating to specific individuals should be treated with respect and care by those who collect them, but also by all other entities in the chain that act upon or conjoin this with other data. (Author/publisher)
Abstract