Implementing Data Governance at Transportation Agencies: Volume 1: Implementation Guide (2024)
Chapter: 2 Why Data Governance?
2. Why Data Governance?
Transportation agencies choose to pursue data governance for a variety of different reasons. This section outlines common factors that motivate transportation agencies to establish data governance structures and processes and highlights some of the benefits that can be realized through data governance.
Goals and Benefits of Data Governance for Transportation Agencies
Maximize value of data to achieve agency goals
Transportation agencies are already making extensive use of data for planning, resource allocation, and development and delivery of projects and services. However, there is tremendous untapped potential to make greater use of data to improve efficiencies and better target resources to achieve agency goals. Building new data reporting and analysis capabilities to get more value from data often requires bringing disparate data sources together. Inevitably, this involves working through a host of consistency and quality issues. Data governance provides the mechanism by which agencies can address data consistency and quality on a systemic basis, which increases efficiency and overall utility for data integration and analysis.
“Data is a precious thing and will last longer than the systems themselves.”
- Tim Berners-Lee
Provide consistent answers to information requests
When an agency executive, legislator or member of the press is seeking an answer to a factual question related to agency activities, projects and services, or expenditures, it is important that they obtain the same answer, regardless of where they go for this information. Data governance processes can designate authoritative sources for different types of information and handle information requests accordingly.
“Data is the center of the universe; applications are ephemeral.”
- The Data Centric Manifesto
Improve data quality and consistency
Poor data quality prevents agency staff from getting value from data and results in lost productivity. It is often the case that people aren’t aware that there are data issues until they try to use a dataset to produce a report or answer a question. At that point, they either give up on using the data or spend precious time correcting data errors. Data governance bodies can improve quality and consistency of agency data by identifying and prioritizing issues, establishing standards, setting expectations for data quality management processes, facilitating collaboration across business units (to address inconsistencies), and supporting tools and training.
“Data that is loved tends to survive.”
- Kurt Bollacker, American computer scientist
Facilitate data sharing and avoid duplication
Transportation agencies want (and in many cases, are obligated) to share data within the agency (across different units), with partner or parent agencies. and with the public. However, agencies face several challenges related to data sharing. First, people may not be aware of what data exists – which can lead to collection or acquisition of duplicate data. Second, the holder of the data may not want to share it due to concern about misuse or
misinterpretation, or fear that it may be used to shed a negative light on some aspects of the unit or agency’s operation. Third, the data may not be in a shareable state due to lack of documentation or poor quality. Fourth, the data may not be easily accessible—it may require extensive manual effort to respond to a data request. Data governance policies and processes can be set up to define what data can and should be shared, how it should be shared, and what steps are needed to prepare it for sharing. Data governance bodies can work to advance tools to enable data sharing.
Common Drivers for Initiating Data Governance
State-level data governance rules or guidance
According to the Beeck Center at Georgetown University, as of 2023, three quarters of the states have established a statewide CDO or equivalent position with responsibility for facilitating data sharing (and in some cases, advancing data analytics) across state government. State CDOs are working with other state agencies to produce data inventories or catalogs, open data portals, and cross agency data sharing protocols. These initiatives have provided the catalyst for state-level transportation agencies to establish data governance leads and programs. Data governance enables DOTs to meet state requirements and support or leverage the state efforts.
Examples of State-Level Data Governance Policy
Colorado Code § 24-37.5-702 (2022) Interdepartmental Data Protocol, Government Data Advisory Board
Connecticut General Statutes Sec. 4-67p – Chief Data Officer. Duties. Designation of agency data officers. State data plan. Agency inventories of data. Open data access plans.
Maryland Executive Order 01.01.2021.09 Chief Data Officer
Michigan Executive Order No 2016-24 Enterprise Information Management
Oregon ORS 276A.353 Chief Data Officer; Oregon Enterprise Information Services Policy 107-004-160
Data breaches
Transportation agencies are at risk for ransomware attacks and other data breaches which are disruptive and costly to the agency and can result in exposure of sensitive data about employees, customers, travelers, or contractors. This risk (or the actual experience of an attack) has motivated some agencies to implement data governance activities focused on documenting and classifying agency data. Data governance processes ensure that sensitive data sources are identified and secured, and that responses to attacks that do occur can be quickly and efficiently focused on restoring the most critical data resources.
Enterprise GIS, Business Intelligence (BI) and data analytics initiatives
Many transportation agencies are building or expanding their enterprise GIS, data reporting and analysis capabilities to increase availability of data for decision-making. There is also growing interest in data democratization – enabling self-serve access to data without the need for technical expertise or IT support. These efforts involve tapping into multiple data sources and creating data warehouses, data lakes, or other central repositories of analysis-ready data. The value of these initiatives depends on having authoritative, standardized, good quality data as well as clear documentation of the meaning of different data elements. Data governance provides the foundation needed for the success of enterprise data improvement initiatives.
Integrating new data sources
Transportation agencies are making increased use of real time data sources including video, roadway sensors, mobile phones, transponders, and connected vehicles. There is a growing
number of private sector data products derived from these sources. New data sources and products are providing new opportunities for insight and efficiency. However, they are also creating a more complex data landscape in which multiple data products may be purchased for similar purposes by different parts of the agency, and data management responsibilities and usage/sharing restrictions are not well understood or integrated into agencies’ practices. This situation has motivated some agencies to put data governance in place, with an emphasis on coordinating data purchases, establishing standard language for purchased or licensed data, and clarifying data roles and responsibilities. Data governance defines processes that coordinate new data acquisition activities to avoid inefficiencies and duplication.
Successful Data Governance Means …
- We distinguish data to be managed as an agency-wide asset based on risks and opportunities to add value.
- We manage access to data in a secure and auditable manner.
- We identifiy, reduce, and avoid creating redundant data.
- We engage the right people in making data-related decisions.
- We are able to resolve differences of opinion on cross-business data issues.
- We have processes to identify common data improvement needs across business units and a way to meet these needs in a coordinated fashion.
- We have assigned roles and expectations for data management and decision-making.
- We have set agency-wide expectations for appropriate behaviors regarding data development and management.
- We provide guidance for strategic agency-wide initiatives to improvie data management.
Data storage
Growth in the volume, variety and velocity of data at DOTs has taxed existing agency conventional data storage resources and resulted in increased use of desktop storage devices, cloud storage solutions and reliance on external contractors. These data storage methods create concerns about ensuring data protection and accessibility. Agencies are seeking more coordinated, managed and economical approaches to meeting changing data storage needs. Data governance works in partnership with IT governance to establish agency-level data storage solutions to meet evolving needs and ensure data security and accessibility.
Supporting digital transformation
Transportation agencies are engaged in the lengthy process of converting paper documents to digital forms. They are also automating a variety of internal and public-facing processes to operate in a paperless fashion. This digital transformation is creating the need for new ways of managing data access and reliability. Data governance defines processes for managing digital data sources including documentation, versioning/archival storage and retrieval, and secure access for authorized users.
Data privacy
There is increasing concern in the US and abroad about collection and use of personal data. Transportation agencies must put processes in place to scrutinize the data they collect or purchase about individuals and their travel patterns. They must justify the need for this data, put safeguards in place to protect it, and communicate how the data are used and protected. Data governance provides the roles and structures for putting data privacy processes in place and enforcing compliance while supporting appropriate types of secure access.
Loss of key data experts
Transportation agencies are facing the loss of long-term employees in a variety of specialized areas, including those that manage key data programs and systems. Many agencies have relied on these seasoned, highly knowledgeable employees to resolve issues and make decisions about the future direction of their programs. When they leave the agency, sometimes there is nobody left with the experience and judgment needed to continue the existing program. Data governance provides a way to embed
best practices for managing data within the organization and reduce dependence on individuals to make sure appropriate actions are taken. It can build a broader understanding of data resources in an agency and develop a wider base of expertise that reduces an agency’s vulnerability when a key employee leaves.
Increase transparency and accountability
Transportation agencies are expected to share information about how they are investing public funds and the processes used to make investment decisions. At a minimum, they must meet state and federal performance reporting requirements. Many agencies have gone beyond this and have established their own performance metrics to provide transparency and accountability – and to drive internal improvement efforts. Implementing effective performance reporting processes requires meaningful and reliable data. It also often requires integrating data across different sources that aren’t necessarily consistent. Agencies view data governance as an essential framework for improving both data quality and data consistency.
Story: How a Lack of Data Governance Impacted a DOTs Ability to Leverage Data to Improve Safety
A failed attempt to implement improved safety analysis. A safety group within a DOT wanted to upgrade their safety analysis methodologies and take advantage of the newest generation of software tools. They wanted to test how much effort it would take to transition their existing data to work within a new analytic tool. FHWA provided technical assistance to help them import the DOT’s existing linear referencing system (LRS) into the database structure required by the new tool. Unfortunately, the import process generated more errors than the tool could handle – due to large numbers of overlapping linear segments. These errors had not been noticed before because prior tools used the LRS in a different way. The lack of systematic data quality checks on the LRS created problems that were only apparent when they were trying to implement the new application.
Lesson: Data governance at this DOT could have established expectations for more rigorous data quality management processes. This would have enabled the safety and GIS groups to identify the LRS issues earlier, avoiding wasted time and effort associated with the failed attempt to implement the new tool. Other business units were being held back by limitations of the LRS as well, but they lacked a cross-functional forum (such as a data governance group) for discussing data issues and collectively pursuing improvements to benefit the agency.
Summary
Transportation agencies face a number of data-related risks, including data loss (due to cyber-attacks, natural disasters and/or lack of backups), release of private or sensitive data, provision of inaccurate data to regulatory agencies, and faulty decisions made based on data that was misused or misunderstood. There are also risks associated with turnover of data experts when critical knowledge about data origins, processing methods, business or reporting requirements, and limitations or interpretations is not documented or passed on.
Establishment of data governance processes together with clearly defined roles and responsibilities can help agencies manage each of these risks.
