BACKGROUND

State departments of transportation (DOTs) have long collected and used data to inform decisions and to manage assets and programs. Much of the data are managed using legacy systems that lack interoperability and are difficult to access and update. These systems continue to be used because they are highly relevant for their intended use and well understood by those who have years of experience using them. However, they may be less useful for decisions that cut across functional areas, involve multiple modes, engage with external partners, or require integration with modern data technologies. In many decision-making contexts, legacy systems can hamper cross-cutting analysis and require extensive investigation of metadata to ensure that the results of an analysis are meaningful for the decision at hand. A common approach to making legacy systems more amenable for cross-functional decisions is to develop a ‘data lake’ or ‘data warehouse’, moving all agency data to a single, enterprise-level platform. While this approach can provide agency-wide access, it does not address differences in data models or the need for a cross-functional, shared understanding of the meaning of the data.

The shift to performance-based management and the need to respond quickly in emergent conditions has made data-driven decision-making imperative for state DOTs. This requires a data governance and management perspective that elevates data as an asset, with the same priority as traditional physical transportation assets. This perspective requires data representations that include data ontologies, glossaries, taxonomies, registries, catalogs, metadata, and models.

Data ontologies are a key element in data representation. A data ontology describes the core information concepts represented in the data that, in turn, drive the business processes and the relationships among the concepts. An ontology ensures not only consistent naming across disparate disciplines, organizations, and information technology systems, but consistent meaning across users. For example, data about a highway bridge would not only include bridge components—superstructure, substructure, and deck—but also traffic volumes carried, noise propagation to nearby receptors, hydrologic conditions below the bridge, stormwater runoff, use of the bridge by birds or bats, maintenance history, construction and maintenance costs, and cultural or historic value of the bridge to a community. Taken together, these interrelated data provide a holistic view of the bridge and its context; a view that can better inform decision-making. A well-designed ontology allows these data to be discovered, integrated, analyzed, and understood by all users without resorting to ad hoc methods that may produce unreliable results and different interpretations. This common understanding of the meaning of the data also supports nimble, multidisciplinary teams that are prepared to collaborate to analyze the data and provide leadership with the best available information to formulate a response.

Research is needed to identify effective practices for developing robust data ontologies and for building agency capacity to use them in transportation decision-making. 

OBJECTIVE

The objective of this project is to develop a guide for state DOTs on strategies for implementing data ontologies that support nimble and efficient data-driven decision-making. 

RESEARCH PLAN

Tasks include:

• Review of current transportation practices in:
  • Developing business-driven data representations (ontologies, data catalogs, business glossaries, etc.); and
  • Processes for using the data for effective decision-making. 

The review will also collect information on agency cultural factors that shape the use and sharing of data across functional areas within a state DOT and with external partners. Cultural factors include workforce strategy, organizational design, capacity for ad hoc teaming, and tools. Information sources for the review include targeted outreach to state DOTs and other transportation agencies (e.g., in-depth interviews), published research, and agency reports.

• Develop case studies of successful practices and lessons learned in integrating data into decision-making.  Case studies may be drawn from public and private sector organizations in transportation and other industries. Of particular interest are case studies of organizations that have transitioned from legacy data systems to modern data-driven decision-making processes.
  
  
• Identify use cases demonstrating the effective use of data ontologies for decisions commonly encountered by state DOTs. Example use cases may include the share of projects delivered on time and within budget; anticipated reductions in crashes from safety improvement projects; or levels of investment in disadvantaged communities.
  
  
• Map pathways for implementing data ontologies and improved processes for data-driven decision-making. The mapping should be a forward-looking assessment of organizational constraints, federal and state regulatory requirements, risks, existing data and data-related processes, and needed organizational and process changes. 

Final deliverables include:

• A guide to successful practices that strengthen the linkages between decision-makers and the data they need to be effective. Topics to be addressed in the guide include: 
  • Successful processes for developing data ontologies that enable shared understanding of the data across functional areas within a state DOT;
  • Identifying needed changes in data management programs and decision-making processes;
  • Sequencing needed changes for effective implementation;
  • Managing a transition from legacy data systems;
  • Building an agency culture for nimble, data-driven decision-making;
  • Case studies and lessons learned;
  • Use cases;
  • Key considerations and the business case for implementing the research; and
  • Other topics from the research. 
    
    
• Stand-alone, customizable communications material (presentation slides, fact sheets, infographics) that summarize the essential concepts of the research and the business case for implementation;
  
  
• A stand-alone Conduct of Research report documenting the project activities; and
  
  
• A stand-alone technical memorandum that identifies implementation pathways, key implementers of the results and well-defined scopes of work for further dissemination and pilot implementation of the research.

STATUS: Research in progress.