Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.
7Â Â C H A P T E R 1 This report proposes a data structure for storing data from bus and rail intelligent transporta- tion systems (ITS), including vehicle locations, passenger counts, and fare collection information. It also describes how the data structure fits into potential transit agency data flow processes that take ITS data provided by vendors, organize and validate it, and use it to calculate key performance indicators (KPIs). As part of that data flow, this report describes tool requirements for critical tools. Tool development, however, is left for follow-up work. In addition to providing a roadmap for data structure usage, data structure definitions, and tool requirements, this report also provides examples and best practices to support transit agen- cies looking to implement the data structure and tools. This section details the background and motivation for this effort and the research process that was followed to inform the development of the data structure and tool requirements. 1.1 Background With the proliferation of automated vehicle location (AVL), automated passenger counters (APCs), and automated fare collection (AFC), transit agencies are collecting increasingly granular data on service performance, ridership, customer behavior, and financial recovery. While granular ITS data can meaningfully improve transit decision-making, transit agencies face many challenges in accessing and using these data sets. Analysts seeking to combine data sets must ensure there is consistency of data identifiers (such as route and stop identifiers) between data from different sources. They must validate data, which includes checking for erroneous and missing values. Data management can also be challenging. Transit agencies must define structures for aggregating, storing, and documenting data in a data repository. Finally, analysts develop and apply techniques for processing the data to provide meaningful information. Each of these tasks requires technical knowledge, staff capacity, and reliable transit ITS data. These challenges are made more difficult by the fact that tools and techniques for validating, linking, and analyzing transit ITS data generally cannot, at this point, be shared across transit agencies because of variation in data collection systems and data formats. Several different vendors provide ITS hardware and software, and data formats vary by vendor. Moreover, agencies may employ a patchwork of ITS that has been acquired and modified over time, leading to further consistency challenges. Standardization of data structures and tools can help address these challenges. Not only can standardization streamline data transfer, validation, and database structuring, it encourages the development of analysis tools that can be used across transit agencies, as has been the case with route and schedule data, standardized in the General Transit Feed Specification (GTFS) format. Introduction
8 Improving Access and Management of Public Transit ITS Data 1.2 Research Process This research employed four different processes to identify the context and need for public transit ITS data management and to collect input from transit agencies, vendors, and other stakeholders, such as researchers and practitioners who have developed and tested other transit data standards. In addition, the Project Panel provided feedback throughout. Figure 2 shows the research process. The following sections summarize the approach for each of these research components. Section 2.1 summarizes findings from the research process. 1.2.1 Literature and Information Review The research team reviewed articles, literature, and web resources to establish a foundation of knowledge on transit ITS data. The review included literature and reports on how transit agen- cies use ITS data, to understand the value of ITS data for transit agencies, as well as their needs. The research team also reviewed existing data standards to identify relevant linkages as well as to understand lessons learned from prior standards-making efforts. The review of existing data standards included: ⢠GTFS. ⢠GTFS-ride. ⢠GTFS Realtime. ⢠Transmodel StandardsâSIRI, OpRA, and NeTEx. ⢠National Transit Database (NTD) reporting standards. ⢠U.S. Department of Transportation (U.S. DOT) Transit ITS Standards Training Modules. ⢠Mobility Data Specification (MDS). ⢠Transit ITS Data Exchange Specification (TIDES). 1.2.2 Transit Agency, Vendor, and Stakeholder Interviews The research team conducted 10 transit agency interviews, five ITS vendor interviews, and five stakeholder interviews. Transit agencies were identified for interviews based on an interest form distributed to transit agencies via several different mailing lists. These included the APTA transit agency list, TRB transit committee lists, and state rural transit agency lists. Ten agencies were selected that represented a variety of sizes, modes, and geographic locations throughout the United States. Figure 2. Research process. Literature and Information Review Transit Agency, Vendor, and Stakeholder Interviews Workshop with Transit Agencies, Vendors and Stakeholders Draft Data Structure and Tool Requirements Testing Workshop with Transit Agencies and Online Feedback Final Data Structure and Tool Requirements Project Panel Input
Introduction 9  Agency interviews included staff from a variety of departments and functions related to ITS data management, analysis, and visualization, as well as those staff who regularly use ITS data for planning, policy development, and operations decision-making. Transit agencies were asked questions about what ITS data they collect, how they manage it, and how they use it. In particular, they were asked about data management challenges and needs, and their attitudes toward data standards. Vendor interviews included two groups: ⢠Data generators, companies that are responsible for hardware that generates raw ITS data. ⢠Data integrators, companies that provide a broad range of services, such as custom data products, and may rely on other vendors to provide data. The research team asked vendors about their data-processing practices and models for pre- paring and transferring data to transit agency clients and about their openness to transit ITS data standards. Stakeholders included researchers involved in using and testing standardized transit data and people who developed or are developing other transit data standards. The research team customized stakeholder interviews for each stakeholder based on their experience with transit data standards. Questions sought to identify important lessons learned and key requirements for data standards. 1.2.3 Initial Workshop with Transit Agencies, Vendors, and Stakeholders The research team hosted a two-part virtual workshop as part of the initial phase of informa- tion gathering. More than 50 people participated in addition to members of the research team. Most participants were transit agency staff, but some vendors and other stakeholders were also included. The workshop provided an overview of the research and an initial proposed approach for developing the data structure and tool requirements. Through polling and interactive small group sessions, the workshop collected information on: ⢠Key use cases for transit ITS data. ⢠How transit ITS data flows from the systems that collect it to the analysts and decision-makers who use it. ⢠Transit data standard adoption roles and processes. ⢠Fields that should be considered in data structure development. ⢠Data transfer and validation tool requirements. 1.2.4 Testing Workshops with Transit Agencies and Online Feedback Process After developing a draft data structure and tool requirements and sharing them with the Project Panel for input, the research team conducted testing workshops with three transit agen- cies. The selected transit agencies varied in size and transit modes, but all three had invested significant effort into their own ITS data management. The objective of the testing workshops was to learn from transit agency staff who had worked closely with this data about whether the proposed data structure and flow could work for their data. Two transit agencies provided sample data to the research team, and the team conducted the following process: ⢠Matched the agencyâs data fields to the proposed data structure, ⢠Converted agency data into the Event Data Files and Supporting Data Files, ⢠Generated the Summary Data Files,
10 Improving Access and Management of Public Transit ITS Data ⢠Reviewed tables for format and data quality, and then ⢠Generated KPIs. In addition to the testing workshops, the research team developed an online document sum- marizing the proposed data structure and tool requirements that could be annotated with com- ments and proposed changes. The document was shared with people who had attended the initial workshop as well as with online communities focused on public transit data. The docu- ment received over 100 comments. Based on the testing workshops and the online feedback, the research team made changes to the data structure and tool requirements.
