Use of Unmanned Aerial Systems for Highway Construction (2022)

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.

20 Case Examples In addition to the questionnaire results, individuals from four state DOTs were interviewed via an online meeting platform to learn more about current UAS applications in highway construction by state DOTs, as well as to identify benefits and potential obstacles state DOTs face when implementing UAS in projects. There were 10 interview questions, which can be found in Appendix C. The following sections describe the characteristics of each project provided by the four state DOTs as case examples and details about UAS data collection and processing, as well as the benefits and lessons learned from using UAS in each of these projects. 4.1 Oregon DOT The Oregon DOT provided information about one of their projects, OR 82: Minam Curve and Bank Stabilization Project, located near Minam in Wallowa County in Oregon, where they used drone data to calculate earthwork quantities. The Oregon DOT conducted the initial flight for UAS demonstration purposes in June 2018 (Figure 12). According to the Oregon DOT, the project office was satisfied with the deliverables (e.g., orthophotos, point clouds, and stylistic visual contents) and requested monthly quantity monitoring. The initial positioning data was not recorded. Consequently, in October 2018, the Oregon DOT located visual markers from the first model and established permanent project controls. Monthly data collection occurred from May to October 2019. The Oregon DOT completed the final quantity measure- ment in July 2020. The final earthwork quantity was based on the material fill site on the west side of the project. The Oregon DOT performed these flights to create a digital terrain model (DTM) to calculate monthly earthwork quantities as well as to calculate the area of rock screening needed on cut faces for this project. They used the earthwork quantities for progress payments and progress monitoring. Figure 13 shows the comparison of the previously collected surface in green, with the current surface colorized by the photos. The volumetric difference in these two surfaces represents the work that was completed between data collection cycles. The red line is a point- to-point measurement to depict scale. The Oregon DOT stated that UAS provided full coverage in a safe and efficient way compared to measuring a sample of surface points using traditional surveying methods. Furthermore, the DOT was able to communicate concerns about completing the project on time by using the UAS data. The Oregon DOT used a commercially available quadcopter UAS for the initial visit. They have established four permanent ground control points (GCPs) using a total station and verified with a GPS unit during each visit. They have also set temporary points (a lathe in an X shape) and measured them with a GPS unit during each visit, as well as new locations as a result of construc- tion activity changing the surface each visit. They used a static laser scanner during an October C H A P T E R   4

Case Examples 21   Figure 12. Overview of the OR 82: Minam Curve and Bank Stabilization Project (Oregon DOT, 2019). Figure 13. Earthwork-quantity monthly comparison (Oregon DOT, 2018). 2018 site visit to provide additional confidence in UAS data. The static laser scanner is a proven survey technology with mature workflows. It provided a control surface to compare UAS data created by the structure from a motion process that was fairly new to Oregon DOT. They used the following specifications in flight planning: 80% forward lap, 70% side lap, all nadir, and single pattern (not crosshatched). They processed the data using Agisoft, a photogrammetry software, and delivered Metashape files to the project office, where volumes were calculated using Bentley InRoads. Oregon DOT stores all data, including the data collected using UAS, on a central server operated by the state. The Oregon DOT’s biggest concern during data collection was about safety. Setting and observing control in an active work area required staff to be constantly aware of their surround- ings. According to Oregon DOT, the actual flights were a big advantage to the traditional survey as they enabled Oregon DOT personnel to operate from a safe distance away from active work. Benefits. According to the Oregon DOT, the project was “extremely successful.” The project office was able to provide timely and accurate data to make payments and address scheduling concerns. The Oregon DOT stated that the additional benefits of using UAS in this project were rock screen quantity calculation and general information photos and videos.

22 Use of Unmanned Aerial Systems for Highway Construction Lessons Learned. Oregon DOT learned that understanding how the data will be used from the start of collection is important. They were able to take their first unconstrained model and do a visual landmark post-constraining. They emphasized the importance of following the best practices for setting GCPs and verify with each visit that they have not moved. The Oregon DOT did not formally document these best practices, but they do exist in the form of field notes, which are a work in progress with every mission that Oregon DOT flies based on experience and which are constantly changing. For this project, Oregon DOT’s approach was to constrain the x, y, and z extents as best as possible. This included setting permanent control points outside of the earthwork area so that it would not be disturbed. The Oregon DOT also established temporary control points based on the benching of the earthwork. The idea was to set points near the break line but make them still visible by the UAS. Between seven and 15 temporary points were set on this project based on the number of benches that were constructed. According to Oregon DOT, for project monitoring efforts, especially during a 24-hour operation, capturing data during the 30-minute shift change was important, so the vehicles and ground conditions were static. The Oregon DOT also emphasized the need to consider different lighting conditions for large-area models (weather and time of day) and found capturing additional stylistic photos or videos to be effective in communicating with stakeholders. 4.2 Ohio DOT The Ohio DOT uses UAS either for collecting as-is data for preconstruction or surveying as-built. In the future, they are planning to expand and use UAS for developing accurate maps. For this synthesis report, they provided information about one of their projects, the US-33 and Pickerington Road Interchange Project, located southeast of Columbus, Ohio, at the intersection of Route 33 and Pickerington Road, where they used UAS to collect as-is data for line work. The data collection was conducted in early January 2021 over 2 days to collect design-level survey data covering 780 acres (Figure 14). Ohio UAS Center personnel used a commercially available quadcopter that was UAS-equipped with an RTK GPS unit for data collection, and the licensed Ohio DOT surveyors shot in the GCPs using their equipment. A total of 50 flights were flown (Figure 15), with each flight being 21 minutes long. The ground sampling distance was 1 cm per pixel. The Ohio DOT UAS flight crew used black-and-white GCPs on areas of grass fields, and road nails and orange paint on asphalt and concrete. The Ohio DOT used a total of 25 traditionally Figure 14. US-33 and Pickerington Road Interchange Project photogrammetric point cloud (Ohio DOT, 2021).

Case Examples 23   surveyed GCPs, and surveyors had checkpoints. The Ohio DOT provided results obtained from the photogrammetry software to the surveyors who then checked the results against the checkpoints. The DOT used district surveyors’ computer-assisted drafting mapping-standard accuracy, 0.07 inches, for this project. Finally, Ohio DOT also considered using UAS lidar for this project but determined that it was not good enough for their own design-level surveying as it creates point clouds that are too fuzzy for design-level accuracies. There is usually a several- inch-wide spread of points on a hard surface (a cross section) in the point cloud, but surveyors need a clean, single line of points. This is difficult to achieve with an airborne lidar unit, even the expensive ones. When done right, however, photogrammetry can achieve good results on hard surfaces. In this project, because of the way the area is set up, they either had to fly over the road or have the survey crew drive to different areas to collect data. Because of the safety concerns about flying over the road, they chose the latter. The Ohio DOT stores raw UAS images on hard drives and processes them at the server rack. Point clouds and orthophotos are sent to the surveyors and stored at the Ohio UAS Center. Benefits. According to the Ohio DOT, UAS “made the whole process much more efficient.” They reported that they were able to collect good imagery and location of data with the UAS at a fraction of the time compared to traditional surveying. They realized great benefits from using UAS in their projects, which saved them effort and time in data collection. They also stated that UAS helps improve safety for the project personnel. Lessons Learned. The Ohio DOT believes that having photogrammetry experts to create accurate point clouds and make accurate measurements is important. They also stated that having surveyors become UAS pilots would be helpful. Finally, they advised that the pilots and the crew involved in data collection should be well trained, as flying drones near the road can be risky. 4.3 New Jersey DOT The New Jersey DOT provided information about one of their projects, Somerville Rail Project, located in Hillsborough Township, New Jersey, where the DOT used a commercially available quadcopter UAS for preconstruction and as-built surveys. The DOT collected UAS data 3 months before construction (Figure 16) for preconstruction survey and on May 28, 2021, Figure 15. US-33 and Pickerington Road Interchange Project photogrammetric point cloud and flight trajectory (Ohio DOT, 2021).

24 Use of Unmanned Aerial Systems for Highway Construction immediately after construction for as-built survey. Even though the UAS was RTK-equipped, the New Jersey DOT did not use this capability. Instead, the New Jersey DOT survey team set up GCPs using traditional survey methods covering a ¾-mile area. The survey team had both the GCPs and survey checkpoints for evaluation. Their major concern was safety during the flight. They believe that the best practice is to postpone the operation if the conditions are not ideal or safe. The New Jersey DOT uses SimpliGov, a cloud-based data storage system custom-built for New Jersey DOT for data storage, including raw drone images and other end products such as point clouds and orthophotos. Benefits. According to the New Jersey DOT, using UAS for data collection has been extremely beneficial because of the safety and precision it provides. Lessons Learned. According to the New Jersey DOT, detailed flight planning is important. They also stated that preparing and having figured out the details of the flight mission can be helpful when contacting FAA for a certificate of waiver or authorization. 4.4 Arizona DOT Arizona DOT provided information for three of their highway construction projects, namely the following: SR-89A Roundabout Construction Project, the I-17 Widening and Improvement Project, and the SR-88 rockslide project, where they used UAS for preconstruction and as-built surveying. The SR-89A Roundabout Construction Project entails enhancements to the roundabout, including signage, striping, and minor pavement rehabilitation to improve efficiency at the site. The roundabout is located at the intersection of SR-89A and SR-179. Arizona DOT flew a commercially available quadcopter UAS in February 2020 for progress tracking and orthoimage documentation. The job site was controlled with six traditionally surveyed GCPs. The project was flown east to west in five flight lines at a flying height of 80 meters, as well as from north to south in five flight lines, also at a flying height of 80 meters for backup imagery. They per- formed all post-processing of aero triangulation, the point cloud, and orthoimages in-house using UASMaster and the Trimble Business Center. Figures 17 and 18 show orthoimages of the SR-89A Roundabout. During UAS data collection, Arizona DOT’s biggest concern was flying over vehicles. To prevent any accidents, the DOT had flaggers controlling traffic, which permitted the flight crew to fly the UAS while traffic was stopped. Figure 16. Somerville Rail Project—preconstruction survey (New Jersey DOT, 2021).

Case Examples 25   Figure 17. SR-89A Roundabout construction orthoimage (Abel Federico and Martin Leveque, Arizona DOT, 2020). Figure 18. SR-89A Roundabout construction orthoimage—GCPs (Abel Federico and Martin Leveque, Arizona DOT, 2020). The I-17 widening and improvement project is located south of Camp Verde, on the I-17 southbound (SB) lanes from milepost (MP) 268 to MP 281. This project was flown in October and November 2020 to develop DTMs. The Arizona DOT used a commercially available quad- copter UAS for the flight mission. The 34 flight missions (each 1⁄3 mile to 1⁄2 mile in length) were flown parallel to SB I-17 at a flying height of 80 meters. Each flight mission was composed of four to six flight lines determined by whether the SB I-17 corridor design was straight (four flight lines) or curved (five to six flight lines needed for coverage). The Arizona DOT controlled the 13-mile job site with 148 traditionally surveyed GCPs depicting the lowest and highest portions of each flight mission. The Arizona DOT performed all post-processing of aero triangulation, the point cloud, and orthoimages in-house using UASMaster and the Trimble Business Center. Figure 19 shows a sample point cloud near MP 280.

26 Use of Unmanned Aerial Systems for Highway Construction The SR-88 rockslide project was near MP 242 on SR-88 near Roosevelt Lake in Arizona. The Arizona DOT collected data in March 2020 using a commercially available quadcopter UAS for documentation and removing boulders and debris and controlled this emergency job site with five GNSS-based AeroPoints GCPs. The project was flown from east to west in four flight lines at a flying height of 80 meters. After reviewing the video imagery and orthoimages, construc- tion crews soon began to push most of the rockslide off of SR-88 and down the slope toward Lake Roosevelt. The Arizona DOT performed all post-processing of aero triangulation and the creation of the point cloud and orthoimage in-house using UASMaster and the Trimble Business Center. Figures 20 and 21 show the point cloud and orthophoto, respectively, of the rockslide near MP 242 on SR-88. The Arizona DOT stored the data in multiple hard drives for redundancy reasons, as they do not currently have infrastructure for saving data to a server or the cloud. Benefits. The Arizona DOT stated that fast, accurate, and safe survey data collection using UAS was advantageous for their agency. One of the major benefits they realized was being able to mobilize the smaller UAS crew as opposed to using traditional aircraft and photo- grammetry cameras. Without a UAS, it would have been impossible to collect data, as heavy equipment had already moved into the area making it hazardous for traditional surveyors on foot. Figure 19. I-17 near MP 280 sample point cloud (Abel Federico and Martin Leveque, Arizona DOT, 2020). Figure 20. SR-88 MP 242 rockslide point cloud (Abel Federico and Martin Leveque, Arizona DOT, 2020).

Case Examples 27   Lessons Learned. According to the Arizona DOT, a lesson learned from prior jobs is to fly projects in multiple directions (if time permits), bring multiple UAS (with different specs useful for different applications), and have a laptop in the field to review imagery before leaving the site. Another important lesson learned from prior projects is the importance of visiting the site in advance to become familiar with surrounding terrain, overhead power lines, shadows, traffic patterns, or suitable locations to launch and land aircraft. Figure 21. SR-88 MP 242 rockslide orthophoto (Abel Federico and Martin Leveque, Arizona DOT, 2020).