Implementing Information Findability Improvements in State Transportation Agencies (2020)

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.

9 The three State DOT pilot projects successfully demonstrated several tools and techniques for improving findability that have applicability in other State DOTs. In addition, the pilots shed light on broader organizational factors that support or inhibit progress in this arena. This chapter highlights key opportunities and challenges identified through discussions with the pilot agen- cies and execution of the pilot tests. These findings may be helpful to DOTs seeking to identify and initiate activities to enhance internal efficiency by making information easier to find. 3.1 Opportunities Several opportunities for improving findability within DOTs are described here. Some of these were demonstrated through the pilots; others were identified through discussions with staff at the participating agencies. All of these opportunities are particularly important to consider not only to meet the needs of internal agency staff but also to address the growing volume of public information requests. Improve Searches Across Agency Policy and Guidance Documents DOTs create and maintain multiple technical manuals and guidance documents. There is often some overlap in content across these different documents. Manuals are typically published as self-contained documents and posted to websites. Opportunities exist to improve the ability to search across the various available guidance documents. Cross-manual searches can provide a broader view of a topic than might be found from consulting a single manual. For example, a search across design, traffic engineering, and maintenance manuals might provide different perspectives on traffic asset selection and placement. In addition, the ability to search across manuals by topic can be helpful for identifying areas of connection to be highlighted, inconsis- tencies to be addressed, and appropriate locations for new guidance to be inserted. Improve Searches Across Construction Projects Practices for managing information about construction projects are typically geared to make it easy to find information for a single project. However, it can be difficult (or impossible) to search for information across multiple projects. For example, a person might want information pertaining to a particular location where several projects have taken place. A person might also want to search for design information produced for projects in similar types of environments or involving particular work types or construction techniques. While some project information management systems include search functions, limited or missing metadata can restrict the usefulness of available search capabilities. In addition, documents pertaining to active projects C H A P T E R 3 Opportunities and Challenges for Improved Findability

10 Implementing Information Findability Improvements in State Transportation Agencies may be stored in separate repositories apart from completed projects, making it difficult to search across all projects. Many project-related documents have a long useful life; improving cross-project search functions gives users the ability, for example, to find deeds or agreements that were created during a project but continue to be in effect long after the project’s completion. Different strategies can be used to provide cross-project search capabilities without creating a new central document repository. The solution piloted at UDOT involved a tool that indexes documents from multiple repositories, allowing people to search these repositories with a single query. Other federated search tools are available that take a single user search query, translate it as needed and execute it within multiple repositories, and then combine the results for presenta- tion to the user. Provide Ability to Filter/Browse by Content Type DOTs produce and store a wide variety of content types. Content types are distinct from file formats (e.g., spreadsheet, image file, and PDF); they describe the nature of the document from a functional standpoint. Sample DOT content types are shown in Figure 4. The ability to search by content type is valuable, since people often seek a type of document rather than a specific known document, for example, “what design exceptions have been granted for this stretch of highway?” Individual content management systems may distinguish content types through controlled vocabulary, naming conventions, folder structures, or metadata tags, but these designations are generally not consistent across repositories. There are opportunities for agencies to create an agreed-upon, shared content type taxonomy. This would make it possible to develop consistent user interfaces for tools that enable precise searching for specific types of content relevant to a task. Application of an authoritative list of content types would allow management system users to quickly find task-specific content, accelerating their performance of tasks, such as conducting a corridor study or creating a Plans, Specifications, and Estimates (PS&E) document for a project. To enable flexibility, the agency content taxonomy can cover high level categories and detail for shared content types (such as correspondence) and allow different business units to build out terms for more specialized content types. Provide Ability to Filter/Browse by Asset DOTs manage information pertaining to a variety of infrastructure assets, including data sets on inventory and condition, inspection reports, test results, guidance, needs reports, and studies. This is an opportunity to implement a “find information by asset” capability, both within individual information repositories and across repositories. Some agencies have created asset taxonomies to support their asset management programs. These asset taxonomies can be repurposed to tag content and create facets for search and navigation. Improve Literature Review Capabilities DOTs with active research programs conduct literature reviews, as part of new project evalu- ation, to see what prior work has been done related to a particular topic. Improving tagging of research reports based on subject area would facilitate these reviews. Tagging based on the Transportation Research Thesaurus (TRT) is helpful but does not go into sufficient detail for DOT-specific topics. Automate Metadata Assignment Full-text searching is powerful and useful, but is not a substitute for the ability to browse, search, or filter content by location, project number, asset, topic, or other key characteristics. For a variety of reasons, files lack the associated metadata needed to enable such advanced

Opportunities and Challenges for Improved Findability 11 search capabilities. Manually creating metadata for a large body of content can be prohibitively time-consuming and may not yield consistent results. Metadata creation can be automated through application of text analytics techniques. Facilitate File Cleanup and Curation Proliferation of redundant, outdated, and trivial (ROT) content occurs in many organiza- tions, including DOTs. ROT content is problematic in many ways: it consumes large quanti- ties of disk space; increases the risk that an outdated, non-authoritative version of a file may be used; and dilutes search results with irrelevant content. The problem is compounded when the individuals who created this content depart the agency; remaining staff may not know which files contain information that needs to be saved. Agencies could use existing tools to identify duplicates and near duplicates. They could also apply text analytics techniques to facilitate the process of identifying which files are “draft” and which files are “final” or authoritative versions. Policies and Procedures Policies Management Directives Engineering Manuals Administrative Manuals Guidance Documents Research Reports Transportation Plans and Studies Long Range Plans Policy and Modal Plans Safety Plans Corridor Plans Local Plans and Studies Land Use Plans County Plans Active Transportation Plans Traffic Studies Transportation Master Plans Transit Plans Project Documents Concept Reports/Scoping Studies Design Memos PS&E Package As-Built Plans Utility Location Plans Agreements Maintenance Agreements Utility Agreements Drainage Agreements Betterment Agreements Right of Way Agreements Environmental Documents Environmental Impact Statements (EIS) Environmental Assessments (EAs) Categorical Exclusions (CEs) Planning and Environmental Linkage Reports (PEL) Environmental Commitments Discipline Reports (water quality, air quality, etc.) Figure 4. Sample DOT content types.

12 Implementing Information Findability Improvements in State Transportation Agencies 3.2 Challenges The most significant challenges that create barriers to improving findability are organiza- tional. All of these challenges were observed during the pilot projects. Fragmented Information Management Responsibilities Adoption of practices to improve findability requires having a person in the agency who can champion and coordinate efforts. This must be someone who (1) views this as part of their role in the agency and (2) has sufficient stature in the agency to commit resources or convince others to commit resources to this effort. The primary challenge of implementing findability improvements in State DOTs is that there is typically no position within the agency charged with the responsibility of making informa- tion findable. One of the pilot DOTs has such a role (WSDOT’s “knowledge strategist”), but this is the exception rather than the rule. In most DOTs, it is difficult to identify a champion who understands the value of findability and advocates for improvement. A DOT library is a logical place to find such an individual, and some DOT librarians have been active and effec- tive in this effort. However, many DOTs have eliminated or scaled back their library functions. Current librarians may not have the necessary bandwidth or authority to get engaged in improving enterprise search capabilities beyond the agency’s library collection. DOTs do have designated staff responsible for various information repositories (e.g., the agency website, engineering content management systems, and records management systems). However, these individuals work within their content silos and don’t necessarily have a holistic, agency-wide perspective. The result is that localized findability improvements at the level of individual systems are possible, but it is difficult to achieve the type of coordinated approach needed to implement cross repository or enterprise search capabilities. Making a Compelling Business Case for Findability Improvements There is no simple or easy strategy for improving findability. Effort is required to implement or enhance search tools, create the metadata needed to facilitate searches, and develop and maintain standardized agency terminology. These improvements, if implemented effectively, can save agency employees time and enable them to re-use existing information instead of re-creating it. Findability improvements can also improve the quality of work by improving access to the information relevant to a particular decision or task. Studies (2, 3) have docu- mented the amount of time employees spend searching for information. The data from these studies can be used to derive estimates of the amount of unproductive search time in DOTs. These estimates can be converted to dollars and used to demonstrate a positive return on investment for improved findability. Another compelling part of the business case for findability improvements is that today’s available tools for internet search and retrieval have set high user expectations. When avail- able workplace search tools are nonexistent or underperforming, they impact employee job satisfaction and the agency’s ability to attract and retain talent. Nevertheless, the business case for investment in findability improvements can be difficult to make. There is more work to be done to better document time savings and quantify benefits of findability improvements in the DOT setting. The importance of information findability resonated for participants in the three pilot proj- ects. They understood the problem and were engaged in initiatives to improve content manage- ment and findability. Even in these agencies, however, participants acknowledged that finding

Opportunities and Challenges for Improved Findability 13 resources for findability improvements was not easy, and the need for such improvements wasn’t widely understood or accepted. Lack of Information Management Discipline If agencies had clear information management procedures that everyone followed, find- ability would be much less of a problem. Organizations that manage findability well have an information architecture that defines where content will be stored and an implemented tax- onomy that describes content consistently throughout the enterprise. Best practices suggest that there should be a designated place to store every type of file, and everything would be in its place, tagged with the metadata needed for future retrieval. Draft and outdated versions of documents would be discarded, and all documents would be text readable to enable full-text searching. In practice, this type of discipline is very difficult to achieve. Agencies that have implemented content management systems have to work hard to get people to use them and supply complete metadata. However, agencies have had success by focusing on high value content types (such as agreements or highway plans), establishing policies and procedures for where and how to store files, providing tools and automated processes that make the pro- cedures relatively painless to follow, and having staff with the time and expertise to facilitate the process. Limited Information Management Competencies Making information findable requires specialized skills and practical experience in build- ing and maintaining tools to facilitate information search and discovery. Relevant skills are taught in library science, data science, and computer science classes. Such skills are in short supply at DOTs. Overcoming the challenge of limited information management competencies at DOTs first involves education about what these competencies are and the value they add. Armed with greater awareness of what is required, DOTs can take steps to build their internal capabilities. DOTs can leverage existing staff, including librarians, business analysts, data analysts, and data scientists to provide some of the needed expertise. These individuals can take advantage of available online and local university training as well as resources of professional organizations to build their understanding of information retrieval techniques, metadata management, taxon- omy development, cataloging/classification, information architecture, usability testing, search engine configuration, and text analytics. Agencies also can use external consultants to supple- ment agency staff, especially for short-duration efforts to build new capabilities. However, it is important that the necessary staff expertise and bandwidth are in place to sustain and continue to enhance the new consultant-developed capabilities.