CHAPTER 7

Overview, Findings, and Future Research

Lidar technology is revolutionizing the transportation industry with its ability to capture highly detailed and accurate 3D data. This synthesis examines the diverse applications and practices of state DOTs in implementing Lidar technology, highlighting the numerous benefits, challenges, and opportunities that accompany its use. Through a comprehensive literature review, online questionnaire, and case examples, this report synthesizes the current state of Lidar utilization, documenting how it enhances transportation planning, construction, and maintenance. It explores current practices, challenges, and opportunities in data collection, processing, mining, and management as well as QA/QC practices and DOT policies and standards.

Overview

Lidar is a technology that is reshaping the transportation industry given its versatility and unprecedented detail. This synthesis explored the diverse practices and experiences of state DOTs in utilizing Lidar to document benefits, challenges, and opportunities. Specifically, the synthesis explores

• The diverse applications and business uses for Lidar;
• Data sources, collection, processing, extraction, storage, and maintenance;
• Data mining;
• Accuracy requirements for applications;
• Data management and governance practices;
• QA/QC practices; and
• DOT policies and standards.

This information was obtained through a detailed literature review, distribution of a questionnaire, and case example interviews.

The pervasiveness of Lidar utilization is underscored by the findings from a questionnaire that achieved a 100% response rate, with participation from 50 state DOTs and the District of Columbia. Notably, almost half of DOTs (49.0%) have used Lidar technology for over a decade to support a wide range of applications. They have seen benefits of safety, efficiency, and data quality from its use. Some have even been able to document strong ROIs associated with Lidar. Others have found innovative ways to share these data effectively across the organization. In contrast, four DOTs (7.8%) indicated they were not currently using Lidar. In those cases, however, the respondent was not aware of the limited usage of Lidar within other departments of their organization.

Despite these many benefits, challenges remain, including limited experience, training, and capabilities; insufficient IT infrastructure; the effort required to extract information; and the need to supplement Lidar data with other sources of information. In many cases, limited budgets

• infrastructure management and development. To date, relatively few NCHRP and DOT projects have focused on geospatial technologies such as Lidar relative to their prolific use and high ROI in transportation applications.

Data Sources, Data Collection, Processing, Extraction, Storage, and Maintenance and Identification of Expected Potential Use of Data

• Research into updating workflows is necessary to truly exploit and leverage the wealth of information provided in Lidar data. In many cases, minimal information is extracted from Lidar data to match the needs of antiquated procedures or standards compared to what is available to be extracted in the data. For example, in the case of extracting profiles, basic elevation information is extracted from a few points on a cross section spaced at distant intervals (tens of meters) rather than leveraging the full cm or dm level resolution available in Lidar data. By transforming workflows, Lidar could provide more robust and comprehensive metrics or analyses that can improve the state of practice.
• Research into effective implementation strategies should consider a DOT’s technology maturity level, current capabilities, and needs. These include (1) exploring training opportunities for personnel in survey and remote sensing divisions as well as across the organization, (2) focusing collaborative efforts between Lidar experts and technical specialists from other domains to tailor outputs that support specific applications, and (3) promoting innovation through collaborative research with academia and industry.
• Emerging technologies such as blue-green bathymetric Lidar and pocket Lidar offer new applications. For example, blue-green Lidar shows promise for hydraulic engineering and bridge scour assessment. Notably, this technology has the potential to improve data collection in turbid water conditions, minimizing the need for manual surveys in dangerous conditions. As pocket Lidar technology becomes more accessible and user-friendly, its applications will continue to expand, providing cost-effective solutions for on-site data collection and analysis.

Data Mining

• Research exploring AI and ML tools for automating data extraction, classification, and mining processes shows promise in reducing manual efforts and improving consistency.
• Collaboration with academic institutions and industry partners can result in the development and refinement of algorithms, enhancing the efficiency and accuracy of Lidar data processing. These partnerships can develop tools to enable extracted information to be simplified to formats that can be integrated into current DOT workflows.
• Additional research is needed to (1) identify the most promising information that can be extracted (as well as that information which AI is not well suited for), (2) develop robust algorithms for extraction of features of interest, (3) compile extensive databases of labeled data from a wide range of sensors, which is necessary to train and validate these algorithms, and (4) explore AI algorithms to infill data gaps to provide more complete information for engineering analysis.

Data Management and Governance Practices

• Research into robust data management practices, including standardized metadata and data retention policies, can result in more efficient Lidar data management. Exploring cloud-based storage solutions could enhance accessibility and streamline the handling of large datasets.
• Integration of Lidar datasets from multiple platforms or other data sources such as SfM/MVS photogrammetry can provide more complete datasets. Research is needed to develop optimal

• strategies for multi-sensor data collection and fusion. Studies on Lidar data sharing should consider the accuracy and applicability of the data to support a variety of business uses within the DOT. High-quality and precise Lidar data can support multiple applications, from planning and environmental management to asset condition assessment, thereby maximizing its utility and fostering better decision-making across different divisions.

Quality Assurance and Quality Control Practices

• Research to develop comprehensive QA/QC protocols can improve the reliability and accuracy of Lidar data. Clearly documenting and communicating data quality and limitations to all users fosters trust and improves the utilization of Lidar datasets.
• Exploration of ongoing training programs for staff is crucial to maintaining proficiency in Lidar technology and data analysis tools. Participation in industry conferences, workshops, and webinars can help staff stay informed about the latest developments and effective strategies in Lidar applications.
• Collaboration efforts between divisions and partner organizations can help establish data quality requirements to support multiple application needs from a single data collection for optimal ROI. Research on determining these needs upfront will promote more effective data and cost sharing. It can also reduce misuse of data collected at a lower quality level for another application or duplication of effort for multiple data collects.
• Improved methods for communicating Lidar data quality and limitations to end users should be explored. Given the wide variability of Lidar data quality as a function of scanning geometry, it is important to move beyond basic reporting of data quality in project reports and metadata, which can only provide aggregate information, toward uncertainty/quality information encoded directly in the point cloud data. For example, Hartzell et al. (2015) developed rigorous pointwise uncertainty propagation models for Lidar data, and O’Banion et al. (2018) developed visualization techniques based on those models to communicate point cloud uncertainty.

DOT Policies and Standards Related to Lidar Data Collection/Maintenance

• Standards, guidelines, and strategic plans to facilitate the adoption, integration, and expansion of Lidar technology within an organization are sorely needed. Regular updates to these standards and plans may reflect technological advancements and evolving organizational needs. A major component of this effort is the transition to 3D workflows, which provides a clear roadmap for leveraging Lidar’s capabilities and ensures alignment with broader transportation goals.
• Research to create a universal transportation data model could facilitate effective data sharing and reuse both within DOTs and with collaborating entities. The development of a consistent taxonomy for extracted assets is important to ensure that data mining can be completed efficiently. This includes a consistent nomenclature and definition for assets as well as more prescriptive details and definitions for defining an asset and its associated attributes. Having a consistent nomenclature will allow information to be shared and integrated across the state DOT as well as with partnering agencies. It can also simplify workflows and data collection efforts, leading to lower costs in the long run. This consistency is increasingly important for improving AI extraction algorithms and supporting autonomous systems that need to operate seamlessly across jurisdictional boundaries.