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.
25Â Â In modern practice, data on asset condition, historical nances, project outcomes, and so forth have become a central and extremely valuable resource for agencies pursuing eective, integrated performance, risk, and asset management. In addition, data allow champions integrating these practices to monitor the progress of their eorts. Data acquisition, management, and governance are oen complex and expensive necessities for transportation agencies. Centralizing and standardizing data governance policies and prac- tices can lead to signicant eciencies by eliminating duplicated eorts and disparities in application. Furthermore, when data are brought together, advanced analytics and business intelligence applications become feasible. is chapter establishes the principles for data governance, along with the associated roles, responsibilities, and documentation observed in practice to be critical to enabling agencies to have cooperative, collaborative discussions and processes (FigureÂ 5). Principles for Data Governance. ese include principles for data assets (the data them- selves), data systems, and artifacts (manuals, guidelines, and publications). Roles and Responsibilities. ese are documented in an asset system catalog (for each system: name, purpose, institutional owner, âsystem-of-recordâ status, technical details, and current access) and an asset data catalog (for each type of asset data: system-of-record and accountable party for creating, curating, and using/maintaining the data). An example of the data life cycle that illustrates these roles is provided below. Documentation. In addition to the data and system catalogs, a DOT should ensure it has technical documentation/user guides for all systems and a data dictionary of the elds/items in each of its databases. All of this should be stored in a centralized location under the super- vision of a designated coordinator. Principles for Data Governance In recognizing the critical role data play in performance, risk, and asset management, the AASHTO Standing Committee on Planning established the following core data principles: â¢ Data are valuable. Data are a core business asset that has value and is managed accordingly. â¢ Data are available. Data are open, accessible, transparent, and shared. Access to data is critical for performing duties and functions. Data must be usable for diverse applications and open to all. â¢ Data are reliable. Data quality and extent are a t for a variety of applications. Data quality is acceptable and meets the needs for which the data are intended. DATA AND SOFTWARE NEEDS
26 Integrating Effective Transportation Performance, Risk, and Asset Management Practices â¢ Data are authorized. Data are secure and compliant with regulations. Data are trustworthy and are safeguarded from unauthorized access, whether malicious, fraudulent, or erroneous. â¢ Data are clear. There is a common vocabulary and definition of data. Data dictionaries are developed, and metadata are established to maximize consistency and transparency of data across systems. â¢ Data are efficient. Data are not duplicated. Data are collected once and used many times for many purposes. â¢ Data are accountable. Decisions maximize the benefit of data. Timely, relevant, high-quality data are essential to maximize the utility of data for decision-making. While perhaps not obvious, the key to efficient integration of performance, risk, and asset management is to follow the principles for data governance. Doing so will greatly ease agenciesâ efforts in beginning and sustaining an integrated management program. Software Central to data management is the software that empowers agency staff to manage, analyze, and understand complex data sets. Transportation networks tend to have a wide variety of asset classes. Pavement condition data may be in linear networks and referencing systems, while bridges, traffic signals, right-of-way, and so forth may be captured in geospatially referenced points or polygons or outside of GIS software in tables. Software packages that perform statistical modeling, complex visualizations, or processing-intensive analyses on these data can be pur- chased piecemeal or as a package with asset-specific modules. To conduct performance, risk, and asset management in an effective, integrated way, DOTs need the following high-level software capabilities. Bridge Management Systems. Ideal bridge management systems perform the following functions: â¢ Upload and retrieve inspection reports and work reports. â¢ Keep inventory and condition at the following levels: â Element [National Bridge Inventory (NBI) Bridge Elements, which include functional, structural, and aesthetic components], â Component (deck, superstructure, substructure, channel, and culvert), â Bridge, â Region, and â Statewide (including an annual report to NBI). Figure 5. Sample data governance map. Recommended Software and Tools ââ âBridge management ââ âPavement management ââ âAsset management ââ âProject prioritization ââ âReporting ââ âSpreadsheet and utility tools ââ âHigh-level language ââ âGeographic information ââ âBusiness intelligence and dashboarding
Data and Software Needs 27Â Â â¢ Model risk-based deterioration of elements. â¢ Model external threats to bridge elements (bridge deck, channel, and culvert) and potential changes in bridge performance goals. â¢ Schedule treatments for elements to minimize the life-cycle cost of bridges. â¢ Generate annual work programs for investment planning and system optimization strategies. â¢ Provide map and visual access to all data. Pavement Management Systems. Pavement management systems support the collection of the measures of cracking, rutting, and raveling that are used to compute both the federal pavement condition metric and whichever other metrics the DOT or local governments have chosen to use. DOTs use them to do the following: â¢ Intake inspection reports and work reports. â¢ Keep inventory of pavement condition by barrel and in short increments (i.e., 0.2 miles or less). â¢ Model external threats to the pavement/roadway prism and potential changes in pavement performance goals. â¢ Support selecting treatments for stretches of road to minimize their life-cycle cost (through risk-based deterioration modeling). â¢ Generate annual work programs for investment planning and system optimization strategies and provide map and visual access to all their data. â¢ Report pavement performance to the Highway Performance Management System. Other Asset Management Systems. Assets beyond bridges and pavement can be managed in GIS-based inventory systems, as modeling and performance reporting for these assets is not federally required. Generally speaking, these systems track the age, location, and work history of assets such as culverts, intelligent transportation systems (ITS) devices, tunnels, signs, sign structures, guardrail, and so forth. Over the past 10 years, DOTs have typically transferred this information from spreadsheets and simplistic databases to GIS software solutions. Project Prioritization Systems. NCHRP Report 806: Guide to CrossÂAsset Resource Allocation and the Impact on Transportation System Performance provides one taxonomy of approaches to project prioritization (Maggiore etÂ al. 2015). It contrasts a âtypicalâ process that relies on assumptions and precedent to direct funding into siloed pools; a âtop-downâ process in which systemwide performance metrics are tied directly to program funding and program sizes are set to optimize the overall set of metrics; and a âbottom-upâ process in which the quantitative benefits and costs of individual projects are weighed and prioritized regardless of program. No matter which of these approaches an agency chooses, it needs systems to support â¢ Generating project proposals (from asset management systems, staff, and local governments), â¢ Distributing funding among programs (using geographic equity formulae or performance), â¢ Prioritizing projects for inclusion in the work program, and â¢ Scheduling the work program. Reporting Systems. Reporting systems may be visualization-based or numbers-based, static or interactive, online or printed. They serve the DOTâs need to â¢ Summarize risk-based asset performance (and other performance metrics) from across an agencyâs asset classes and functions; â¢ Present performance metrics required by state and federal rules; â¢ Justify investment in the transportation network by clearly communicating benefits, appropriate choices in trade-offs, and negative consequences of underinvestment or disinvestment; and â¢ Present a modern, forward-thinking, friendly, innovative face for the DOT to legislators, stakeholders, and the public.
Data and Software Needs 29Â Â When software tools to support agency operations related to performance, risk, and asset management are being identified, it is important to carefully consider all options. The various strengths and weaknesses of alternatives should be weighed, including how well a given soft- ware option suits a specific task, how accessible the option is (e.g., what training would be required to utilize it fully), how well the software scales for various deployment levels, and the overall cost of the software, as well as how it might support the broader vision of integration. Data Implementation To support an integrated management approach, it is often necessary to acquire, manage, analyze, and visualize large amounts of data related to asset condition and location, financial planning, risk probability and impact, and more. This may require additional recruitment or training of technical staff. Additionally, with competencies in specialized dashboarding and business intelligence solutions, agencies can produce highly interactive and effective desk- top or web-based dashboards with dynamic visualizations that can be especially effective for decision-makers who need to explore distilled data sets in real time. Accessing and Analyzing Data Data science is becoming increasingly common in modern transportation practices as new data, software, and technologies emerge. To leverage these valuable resources, agencies must be able to identify and implement those practices that can best support their public obligations. Such resources may include detailed infrastructure inventory data from lidar collection; live, crowdsourced infrastructure condition or incident data from smartphone applications like Waze; or intricate road user behavior data from cellular providers or insurance companies. Such resources require advanced competencies in data acquisition, management, and manipulation through programming software such as SQL, Python, and source-specific application program- ming interfaces. To effectively make use of advanced data sets, agencies must also have the capacity to properly analyze and distill complex information. This may require the use of high-powered statistical software such as SAS, R, or Python to programmatically operate over extensive, dynamic data. This software may also be used to develop models to help solve problems related to missing data, asset condition forecasting, or financial planning approaches. Communicating Findings Once data have been processed and analyzed, it is critical that an agencyâs technical staff can effectively communicate results to their colleagues, agency leadership, and the public. This is commonly a point in the technical process where a great deal of information has the potential to get lost in translation. No matter the quality of the underlying analysis, manipulation, and modeling processes, if proper communication of results is not achieved through appropriate reporting and visualization, agency decision-makers will miss out on critical information. As data reporting and visualization are highly customizable and come in many forms and levels of complexity, agencies should continuously work to advance these practices through training and staff development programs. Similarly, the technology and available software to support these practices are constantly evolving, and this is leading to more effective and acces- sible options. Common programming skill sets of Java and Python can support programmatic creation of highly flexible reporting and visualization products. With the geographic nature of transportation assets, mapping and geospatial visualizations of asset data such as condition, VTransâ Centralized Data Strategy VTrans implemented a centralized GIS-based data system to support data access across discipline backgrounds. This system includes the Vermont Asset Management Information System.
30 Integrating Effective Transportation Performance, Risk, and Asset Management Practices operations, traffic safety, and more can be highly effective. This is achievable through licensed GIS software such as ArcGIS or Tableau or through open-source alternatives such as QGIS or the GeoPandas and related packages in Python. Maintaining Credibility While emerging sources of data can provide helpful insight to planners and engineers at transportation agencies, they can also lead to overconfidence. FHWA rules require performance targets to be set and updated every 2Â years. Often, these targets are set using inventory data, performance models, and software packages that are not fully understood or explicable by agency staff, perhaps because they are off-the-shelf packages, because they are based on national data sets and parameters, or because the agencies lack the skill sets to manage the tools and have outsourced the task to consultants. In that environment, the data and software themselves become risks. If the agency cannot explain its models, its projections, its targets, and why it may meet or miss those targets, it may stand to lose funding flexibility (i.e., the FHWA âpenaltyâ) and credibility with executives and legislators. Data and software implementation cannot be as simple as procuring a package or contractor and reporting a number from a black box. Agencies should fully understand every parameter in their performance models and the impact those parameters have on results. With that under- standing, they should be able to build models tailored for their agency that they can confidently defend. Using sensitivity analysis, they should able to explain the range of possible outcomes on the basis of realistic variation in external factors. When reality fails to match their projec- tions, they should be prepared to explain why this occurred and to refine anything that needs refinement. Agencies may also wish to reconsider the value of models or software whose complexity exceeds the capability of the staff, either because of a lack of technical skills or because of a lack of time/bandwidth to learn the details. A simple linear model may have the same predictive value as a complex multivariate approach in a high-level analysis with uncertain parameters, and the linear model will be much easier to explain. Modern and Emerging Data Sources As software and technology continue to progress at a rapid pace, new and powerful data sets are becoming available to agency practitioners. These modern resources are beginning to allow for all-new, highly effective, and novel solutions to existing tasks encountered at all levels within transportation departments. Efficient acquisition and implementation of such data sets through well-designed frameworks can make way for the integration of performance, risk, and asset management. This can be achieved by providing a revolutionary look into the details of network performance and limitations, motorist behaviors, and subtle underlying risks within infrastructure systems. Primary categories of emerging data sources include connected travelers, connected vehicles, connected infrastructure, and other connected sources. Critical to utilizing these data sources is recognizing how each can be consumed, such as what their points of access will be and how acquisition, marshaling, and analysis can practically be performed. There are many extremely valuable new data resources being made available as technology progresses, and it is critical that agencies begin to develop effective and, equally important, efficient processes for acquir- ing, managing, and implementing them. A broad variety of tools and solutions is being made
Data and Software Needs 31Â Â available through open-source resources as well as modern industry enterprise solutions from major companies. Agencies should consider the relative cost of such options and the long-term implications related to adoption and evolution of technology; flexibility and ease of use; as well as security, veracity, and more. The Intelligent Transportation Systems Joint Program Office of the U.S. Department of Transportation has developed a series of reports titled âIntegrating Emerging Data Sources into Operational Practiceâ that discuss in extensive detail the state of the practice in big data tools and technologies and their applications for transportation agency practices (Gettman etÂ al. 2016, 2017; Sumner etÂ al. 2018). Additionally, ongoing research that is investigating tools, methods, and guidance for managing emerging transportation data and technologies and leveraging them for advanced agency decision-making is taking place under NCHRP Project 08-119, âData Integration, Sharing, and Management for Transportation Planning and Traffic Operations.â NCHRP Project 08-116, âFramework for Managing Data from Emerging Transportation Technologies to Support Decision-Making,â provides guidance to help agencies initiate a shift away from traditional data management practices to a practice that is more effective in handling modern big data sets from emerging technologies. The steps in the big data roadmap detailed and illustrated in NCHRP Research Report 952 (Pecheux etÂ al. 2020) are as follows: 1. Develop an understanding of big data. 2. Identify a use case and an associated pilot project. 3. Secure buy-in for the pilot project from at least one person in leadership. 4. Establish an embryotic big data test environment. 5. Develop the pilot project within the big data test environment/playground. 6. Demonstrate the value of the data to other business units. 7. Demonstrate the value of the data to executives. 8. Establish a formal data storage and management environment.
32 Integrating Effective Transportation Performance, Risk, and Asset Management Practices Integration Maturity: Data and Software Needs To assess an agencyâs level of maturity in the key topic of data and software needs, the bench- marks below may be considered. 0 2 3 Level Preintegration No actions have been taken to pursue integration within the agency. Data management practices and data sources are largely siloed, with no active interest on the part of agency executives in pursuing the integration of those used for performance, risk, and asset management and no notable integration of data practices or sharing of key data resources. Level 1 Initial No practical changes have been made to the operations of th eagency or the acquisition, management, or use of data or software systems, but agency decision-makers are discussing the potential costs and benefits of formalizing and modernizing their architecture. Level Level Level 4 Defined The agency has begun to develop a roadmap for modernizing data collection and governance as well as performance, risk, and asset modeling. Staff may be exploring new statistical modeling methodologies, visualization techniques, or software solutions. New, tentative lines of communication are opening between departments about the potential for data sharing and the reduction of duplication of effort. Expandable, Repeatable Integration has begun to move from theory to actual practice. Agency processes are being transformed according to the defined integration roadmap and are becoming systematic and repeatable. The effort is beginning to produce some tangible outcomes, though processes are still changing rapidly as integration expands through departments and affects day-to-day operations throughout the agency. At this point, integration has major implications for staffing and resource needs. This stage may last for an extended period as integration permeates the agency at various levels and departments. Managed Agency data and software processes are being transformed according to the defined roadmap, with new data and workload sharing becoming systematic and repeatable. A formal data governance structure is being established by using enterprise systems to minimize duplication of effort, improve efficiency of data collection, and increase availability of key data and software resources for agency staff across multiple departments. Some advanced data use practices are being adopted, including statistical modeling, geospatial and other visualizations, and linkage of key data sets. At this stage, the agency may be increasing investments in data and software systems without experiencing immediate financial returns; however, new structures and practices are expected to form the basis for major efficiencies. Evolved practices may also precipitate the need for additional technical staff or training for existing staff to achieve competencies needed for the new data and software systems. Level 5 Optimizing Efficient, aligned data collection, governance, and management are the standard operating procedure at the agency, and management pursues continuous improvement. Collection of field data efficiently feeds information to management systems fitted with vetted, trusted models for asset performance that identify treatment plans for the full asset life cycle and identify risk. The benefits of investment in these assets are clearly communicated to decision-makers, legislators, stakeholders, and the public through a dashboard and visualizations.
Data and Software Needs 33Â Â Vermont Asset Management Information System (VTrans) VTrans, Vermontâs department of transportation, recently undertook a comprehensive review of its data resources and management systems to support asset management as part of the development of the Vermont Asset Management Information System (VAMIS). The VAMIS puts many of the data governance principles included in this section into practice and is an effective foundation as the development of integration efforts at VTrans matures. A detailed internal document has been developed to define management systems to be included in an enterprise data environment along with a detailed list of the categories of data, their functions in supporting VTransâ business processes, and individual staff roles in data stewardship and use. Additionally, VTransâ Survey123 GIS tool is being used for postdisaster data collection. The system incorporates drop-down menus and has simple user interfaces on tablets for field engineers to use to input data. VAMIS will help connect Detailed Damage Inspection Reports (DDIRs) and project cost info to the Survey123 tool. VTransâ Construction Management System also is being linked to tie money down to the asset level. This tool will improve VTransâ risk management capabilities and make the stateâs transportation system more resilient. It will provide better insight into the causes and effects of damage for planning purposes, improving both long-term asset management and short-term operational decisions able to be integrated into performance and asset management efforts. VTrans pulls data needed to produce annual performance reports like VTransparency and the VTrans Fact Book from across the agency. VTrans is moving toward online information management systems that make data available and visible throughout the organization. Examples of key data sources and information systems include the following: DDIRs with information about storm-related damage and other events that cause damage and disruption to the stateâs transportation system; Linear Reference System with roadway inventory and cross section data; VAMIS and the legacy Pavement Management System and Bridge Management System for asset-level data and the asset inventory (location, condition, treatment history, and inspection reports); Construction Management System for things like project cost and schedule information; Managing Assets for Transportation Systems (MATS), which records most highway maintenance work by location (MATS is being expanded to include culverts, transportation buildings, and ancillary assets); Airport Information Management System, which identifies, prioritizes, and tracks progress on aviation-related projects (the primary project driver at both the federal and state levels is aviation safety); and Corridor Needs Tool, an ESRI StoryMap that visualizes geospatial information about asset, community safety, and traffic needs. VTrans is also incorporating actions related to data governance in its Asset Management Implementation Roadmap. As part of initial steps in development of VAMIS, VTrans staff and staff from the Enterprise Project Management Office (EPMO), a division of the Agency of Digital Services) are connecting business needs with functional and technical requirements for the new system. The EPMO staff conducted interviews with VTrans staff that culminated the development of a context diagram, which shows how information would flow into and out of the VAMIS in support of key functions of VTrans. While this effort was limited to a single information system, conversations about automating data validation, quality assurance improvement needs, and connecting data to processes and agency-level objectives will benefit future information technology development and deployment initiatives. Integration in Practice