States are using data from unmanned aerial systems to help predict geological threats, prioritize mitigation efforts, and aid recovery after an event occurs—with reduced costs and improved safety.

Transportation projects use unmanned aerial systems (UAS)—often referred to as drones—for many purposes, such as aerial photography, reconnaissance, surveying, structural inspection, and monitoring for documentation, safety, and security. The benefits of using UAS include improved access, better quality, increased safety, greater speed, and improved efficiency.

Unmanned aerial systems (UAS) can provide geotechnical data to transportation agencies while increasing safety and efficiency. These systems can obtain important information, such as rock joint condition and orientation, in locations with difficult access.

Although most geotechnical information lies below the ground surface, UAS can evaluate landslides, rockfall, embankment distress, settlement, sinkholes, and similar conditions where measurements associated with visible ground features and ground deformation provide useful information about the subsurface. UAS can also detect the rate of change of these features—which can offer insights on causality, such as seasonal or weather influences, like rainfall events.

UAS are most productive in dangerous or hard-to-access sites, or locations where an aerial big picture bird"s-eye—or perhaps "drone"s-eye"—view is useful, or critical, to a complete understanding of a site. Aerial views are often beneficial because of the large scale of some distress features associated with landslides and rockfalls. For geotechnical purposes, high-resolution photos, three-dimensional (3D) point measurements from airborne laser scanning (commonly LiDAR, light detecting and ranging), or structure-from-motion image processing and associated topographic change detection are powerful tools to characterize sites, assess defining geologic features and geohazard threats, and measure movement over time.

A rockfall event at the southern portal of the Silver Creek Cliff Tunnel on Minnesota Highway 61 in November 2018 filled the catchment area between the rock face and the protective barrier. UAS can provide safer inspection of sites with recent failures, particularly when site safety is further reduced by snow and ice cover.

The Impact of Rock-Slope Failures and Rockfalls

The American West is known for its rugged geography and impressive mountainous terrain. Whether driving on I–70 near Idaho Springs, CO; traveling on the Seward Highway between Anchorage and the Kenai Peninsula in Alaska; or approaching the Knapps Hill Tunnel on U.S. 97A in Washington State, drivers may encounter an extensive delay or road closure due to a rockfall event. These events are not limited to the West. Many examples of soil and rock slope failures also come from the eastern United States, from the Blue Ridge Mountains in North Carolina to the White Mountains of New Hampshire, or even mid-continent areas along rivers and lakeshores, such as in northern Minnesota. While geologists and engineers strive to design safe highway corridors, rockfall events remain a regular occurrence from of a variety of causes. Natural weathering, freeze-thaw cycles, vegetation, earthquakes, rainfall, and other conditions can lead to rockslides and rockfall.

Rockfall events often result in significant impact to the traveling public. Aside from the obvious safety concerns posed by large rocks falling near or on vehicles, drivers avoiding these rocks or maneuvering to reroute can create additional hazards. Roadway closures to prevent crashes and ensure motorist safety, which often occur with little or no advance warning, can result in significant inconvenience from initial travel delay and lengthy detour routes. Large-scale rock-slope failures can even disrupt routes for months and impact local and regional economies, as described in a 2010 report, Economic Impact of Rockslides in Tennessee and North Carolina, prepared for the Appalachian Regional Commission.

Rockfalls can be controlled by a variety of proactive measures that include rock scaling, trim-blasting, mechanical stabilization, and rockfall catchment systems. These are often effective strategies to reduce the risk to the traveling public. Some sites use multiple control techniques. Proactively preventing all possible rockfall events is impractical, but tools are available to help prioritize mitigation and aid recovery after an event occurs.

Minnesota Highway 61 at Mt. Josephine is the site of two rock cuts where rockfall reaches the travel lanes. The Minnesota Department of Transportation completed the initial surveying with a total station—ground-based survey equipment—followed later with a terrestrial LiDAR survey and point cloud photogrammetry using a UAS.

The success of the pre-event risk assessment or post-event mitigation strategies is based on an understanding of failure mechanisms, geometry, and the condition of the materials. While the strength of intact rock is an important characteristic, the discontinuities within the rock mass structure—such as the fissures, fractures, joints, faults, and bedding planes, and their condition— usually control how failures occur. These features often follow a geometric pattern locally and sometimes over large areas such as complex joint sets and systems.

Obtaining quality rock slope information can be a risky endeavor involving putting people on potentially unstable terrain to map out geometric features and characteristics of rock formations. Today, transportation agencies are using UAS to conduct these surveys more safely and more efficiently. The digital information acquired by UAS can be quickly processed and deployed to help expedite project delivery.

For example, the Colorado Department of Transportation (CDOT) created a digital model of a rock slope using UAS data and quickly designed a trim-blast after a failure closed I–70 near Dumont, CO, in November 2019. Representatives from CDOT and its contractor viewed the model in a video-conferencing meeting to collaboratively identify and delineate areas of the rock slope for removal. CDOT then shared the model with the blasting company to use for the blast design.

Return on Investment

Many State departments of transportation are investing in geohazard mapping and management programs, as well as the tools to provide the input data for processing, review, modeling, and decisionmaking. UAS is a new option in the geotechnical toolkit that is proving highly effective for improving the quality of information, increasing safety, and reducing project cost, risk, and effort. With better understanding and active geohazard management to anticipate events and proactively plan activities, transportation agencies can efficiently apply rock scaling and removal or other strategies, plan road closures and detours, and inform the public of potential impacts with advance notice.

Technology of various forms, including UAS, is making geotechnical and geohazard asset management easier, safer, faster, and more economical. A short UAS flight in the field can provide data more comprehensively than would have been previously provided by a team of two to three qualified geologists or geo-engineers hiking in an area and manually mapping for several days. UAS can be outfitted with several types of imaging sensors for high-quality photos, structure-from-motion 3D point measurements, and airborne laser scanning.

Topographic change detection, which can be performed more frequently and on a smaller scale with UAS, is a transformative area of development. This process involves comparing two or more temporal datasets. Current applications include larger landform and land use changes (such as open pit mining), mapping changes in snow cover and glaciers, and avalanche detection and mapping. There is a significant return on investment for geologists, geo-engineers, agencies, contractors, the traveling public, and taxpayers.

CDOT is using UAS to collect baseline images after most rockslides. At small sites, the cost of aerial data collection from UAS ranges from 10 percent to 20 percent of the cost of using a helicopter for similar data, depending on location. That cost flips when attempting to collect corridor scans: the time to collect images of a corridor (anywhere from 5 to 10 miles [8 to 16 kilometers] of highway slopes) results in a cost about 200 percent to 500 percent of the cost of collecting images using a helicopter. However, in CDOT’s experience, the accuracy of GPS data in images taken from a helicopter has been a concern, and the quality of data is appreciably better from UAS in both image quality and in the ability to process the images.

While CDOT does not have good cost data that show the value of the better quality of the drone image compared to the photographic images from a helicopter, the agency recognizes that the models are much better with UAS-collected images. For CDOT, this accuracy has made UAS worth the additional cost when also considering the value of the model quality—a model created from helicopter-collected images may not provide the opportunity to see change detection at a desired precision for analysis.

An aerial view of a rock slope, and an adjacent highway and rail line, as seen from a UAS used on a New Hampshire Department of Transportation research project examining how UAS systems can increase safety and decrease costs of transportation projects.

Geotechnical UAS Operations

State DOTs generally have established an organizational structure to facilitate safe and efficient UAS operations. Like other UAS operators in both the public and private sectors, geotechnical UAS users must adhere to statutory and regulatory requirements. Application limitations include flight time, location, weather (high winds or inclement conditions), stray currents, and magnetic or radio frequency interference. There could also be poor global positional control, depending on terrain, which would require manual flight by the UAS operator.

A key consideration in the specific use case of rock slope assessment, given the steep and highly variable terrain, is that UAS, including flight planning software, needs to have terrain-following capabilities. In order to have high-quality data from LiDAR and photogrammetry, the ground sampling distance to the sensor must be held constant. Some UAS software can only enable flight at a fixed elevation, rather than at a fixed distance to ground, which is critically important not just for data collection but for safety of UAS in environments where the terrain can change rapidly.

Data can be captured manually or at predetermined intervals. Images taken by UAS are transmitted back to the remote controller, which is typically connected to a mobile data collection device for live image processing. Once the images have been downloaded from the UAS, an extensive array of software is available to process the data for different purposes, including the following:

Photogrammetry, digital aerial photography, and structure from motion—

Measurements from UAS photographs combined with correct GPS coordinates produce extremely accurate mapping. The flight images obtained from the UAS overlap, and algorithms within the image processing software can identify related features in each image that enables the images to be stitched together. Dense point clouds can be extracted from digital aerial images, and the density of the data can be similar to airborne laser scanning systems.

Digital terrain modeling—

Computer algorithms can predict terrain while ignoring vegetation; digital elevation modeling operates similarly but includes plant life. Both types of modeling enable accurate contour mapping of features.

Airborne laser scanning and 3D modeling—

Points are processed using computer software to form a lifelike image composed of a 3D reality mesh model. Analysts can rotate the images produced by the software to provide improved site understanding, to help visualize potential issues, and to identify surface anomalies in the context of the surrounding elements. The 3D models differ from photos in that they are constructed of 3D information that can be employed as a framework to show relationships with other spatial data.

The high-quality digital photos obtained from UAS, even without the benefits of processing, are useful to observe sites from otherwise inaccessible vantage points. For large events, UAS observations often offer a more complete site assessment, providing a larger field of view than is typically available on the ground. This is particularly advantageous when capturing the size and severity of the problem is important.

Site overview imagery is often invaluable for showing nontechnical observers the scope of a landslide, like this one in North Carolina, or rockfall event beyond simple cleanup of the travel lanes.

Managing Geohazards in Colorado

CDOT’s Geohazard Program has used some form of aerial data collection since the 1990s. At that time, photographic images mainly documented construction activities and reviewed oblique, aerial images of steep slopes. Recently, the program has been collecting baseline images, in the form of structure-from-motion and LiDAR point clouds, of geohazard corridors and after geohazard events.

The University of Vermont Spatial Analysis Lab produced this 3D point cloud from rock slope inspection data for the New Hampshire Department of Transportation. The work was done as part of the study on UAS applications to reduce cost and increase safety for traffic operations at Crawford Notch State Park.

“With these advances in aerial data collection and processing over the past several years, the use of UAS has become part of the workflow for several of our programs at CDOT,” says Stephen Harelson, P.E., chief engineer at CDOT.

Over time, as additional data are collected, change detection will direct investigation to specific slopes within in a corridor. In combination with other information such as UAS rockfall hazard rating system, precipitation, temperature data, and change detection maps created from surveys, the agency’s goal is to establish the likelihood of an event and manage risk. Based on an informed probability, CDOT can direct detailed investigations to those areas of a slope that appear most susceptible to failure based on data and subject matter expert observations and review. It is important to note that even the most informed tools and techniques still only provide predictions of future outcomes.

Geotechnical Applications in North Carolina

For emergency landslide roadway closures, the North Carolina Department of Transportation (NCDOT) Geotechnical Engineering Unit supplies a working plan within 24 hours, typically with biddable items provided within 72 hours. UAS has facilitated the entire process by enhancing four critical general project items:

• Fast aerial overview to provide NCDOT management and the public with an idea of the scale and potential impact of the problem.
• Supplemental information from the field review to develop the immediate mitigation approach, scope, specialty geotechnical repair items, quantities, and mitigation limits (for potential right-of-way and permitting requests and requirements). This includes information necessary to assess slope stability and the potential for partial highway reopening prior to project completion.
• Periodic overview of progress during mitigation to inform management, the public, and project inspectors without requiring hazardous slope access.
• Initial and final orthophoto, topographic, and point cloud projections, requiring limited hazardous surveying and estimating, for final project pay quantities.

NCDOT used UAS on a project along I—40 in February 2019. A simple initial overview provided general quantities, mitigation boundaries, and specialty items anticipated from the contractor. The UAS images helped observe and document project progress and assisted in determining when conditions were at the point where rockfall barriers could be placed and detour travel lanes opened.

Derivative products from UAS orthoimagery (small-scale photogrammetry) enable transportation agencies to assess earthwork pay quantities. The outlined area illustrates the location where slope mesh and retaining anchors will be installed. NCDOT and its contractor agreed on this method of measurement for payment, decreasing the work required for onsite inspectors.

“Aside from the established and emerging technical applications of UAS data,” says D. Clayton Elliott, a geological engineer in NCDOT’s Western Regional Office, “the ability to provide an immediate oblique view of many geotechnical projects for concept, safety, and construction is now an expected part of any project.”

New Applications Evaluated in New Hampshire

The New Hampshire Department of Transportation (NHDOT) recognized that the use of UAS had the potential to reduce costs and increase safety for a variety of transportation operations. In 2017, NHDOT, in partnership with the University of Vermont’s UAS team, began a research project focused on evaluating UAS technology for a broad range of case studies including rock slope inspection. One case study focused on the inspection of a rock slope near Crawford Notch State Park in July 2017. The University of Vermont’s UAS team conducted three flights, collecting 310 photos during a total flight time of just over half an hour. The effort had two goals: (1) create a high-resolution georeferenced point cloud of the rock slope suitable for 3D modeling to analyze the rock structure, and (2) capture high-resolution inspection photos of the rock slope to provide multiple viewpoints of the rock face.

The team created a 3D model using the 310 images and digital photogrammetric techniques and generated a seamless orthorectified image mosaic. The processing for this project took approximately 70 minutes to complete. The final report, The Integration of Unmanned Aircraft Systems to Increase Safety and Decrease Costs of Transportation Related Projects and Related Tasks (FHWA-NH-RD-26962J), provided several findings related to inspection and safety. First, the UAS provided a view of the rock slope that an inspector would be unable to view from the ground. Second, working on the ground near rock slopes is not risk-free because of potential rock fall and frequent roadway traffic. Using UAS generally keeps personnel safer from potentially dangerous rock slope site conditions. Third, the detailed 3D rock slope models provided measurements in locations that are unreachable by manual measurements, or only accessible using potentially dangerous rope-access methods.

Integrating Surface and Subsurface Measurements

Surface mapping and imagery from UAS can add even greater project value when used in combination with other geotechnical tools, such as a borehole televiewer. Optical and acoustic televiewers can be employed in boreholes to gather high-quality information on subsurface stratigraphy as well as rock joint and fracture orientation.

The Minnesota Department of Transportation (MnDOT) relocated a portion of U.S. 53 between Eveleth and Virginia, MN. Future iron ore mining adjacent to the alignment has the potential to create excavated rock slopes up to 500 feet (150 meters) high, in addition to existing slopes, requiring characterization of rock discontinuities. MnDOT used two innovative methods to collect the information: down-the-hole televiewer logging (both optical and acoustic) and photogrammetry using images acquired via UAS.

Once the team completed the field work, they began the joint mapping task. A private consulting group, one of MnDOT’s project partners, first constructed the models from photographs and geo-referenced them using a structure-from-motion application and contact-free measurement of geological/geotechnical parameters. The 3D images provided comprehensive documentation by reconstructing the geometry of the rock walls, with measurement of geometric and geologic features represented as points, distances, areas, and orientations. The optical and acoustic logging was most useful for determining the spacing between recurring joint-sets (persistence), and the photogrammetry was only helpful for estimating persistence. During construction excavation, the persistence was shown to be greater than measured using photogrammetry because the full extent of joints was obscured by talus and vegetation.

“[We have] found UAS imagery, plus the resulting point clouds and meshes, to be invaluable in characterizing inaccessible rock exposures,” says Dr. Lee Petersen, P.E., a principal engineer with one of MnDOT’s partners. “We typically capture still photographs to support photogrammetry (to make the point clouds and meshes), and videos to better understand rock outcrop geometry and discontinuities. The point clouds are used to extract discontinuity location and orientation, and meshes are used to extract rock block geometry.”

An example image of mapped rock joints from a UAS photogrammetry survey along a rock slope at a proposed bridge location on the U.S. 53 relocation project in Virginia, MN.

Advanced Geotechnical Methods in Exploration (A-GaME), an initiative of round 5 of the Federal Highway Administration’s Every Day Counts (EDC) initiative, listed televiewers and UAS as highlighted technologies (www.fhwa.dot.gov/innovation/everydaycounts/edc_5/geotech_methods.cfm).

Integrating UAS into Asset and Management Programs

UAS provide another tool in the geotechnical toolbox for site characterization. Benefits include fast response to emergency events, decreased data acquisition time, reduced cost, greater operational efficiency, improved quality, and significantly improved safety at hazardous or unstable sites. Transportation agencies’ use of UAS has increased significantly in the past few years, and many State DOTs have created new programs related to UAS applications. The photos, videos, and 3D measurements obtained from UAS are already improving standards of practice for geohazard visualization, change detection, risk assessment, informed decisionmaking, and hazard/asset management.

UAS use increases productivity during data collection and enables more efficient processes and more rapid detection of areas of concern. These advantages, combined with generating high-quality end products, are significant benefits considering the additional value that total data collection costs may decrease due to reductions in labor, time, and other expenses.

UAS technology, coupled with geotechnical project implementation, provides an opportunity for improved practice merging surface and subsurface data for a significant return on investment and improved risk management—leading to reduced costs, improved safety, and, perhaps most noticeable to the public, fewer traveler delays because of unexpected road closures from geohazards.

Derrick Dasenbrock, P.E., D.GE, is a geotechnical engineer in FHWA’s Resource Center, joining FHWA in 2020 after working in MnDOT’s geotechnical engineering section.

James Gray, P.E., is a UAS and construction technology engineer in FHWA’s Office of Infrastructure. He leads FHWA’s EDC-5 Unmanned Aerial Systems innovation.

Ben Rivers, P.E., is a senior geotechnical engineer in FHWA’s Office of Technical Services. He leads FHWA’s EDC-5 A-GaME innovation.

Ty Ortiz, P.E., is the geohazards program manager for CDOT.

Jody Kuhne, P.G., P.E., is a regional geological engineer in NCDOT’s Geotechnical Engineering Unit (GEU). He is a 27-year employee of the NCDOT GEU, based out of Asheville, NC.

Krystle Pelham is an engineering geologist in the Geotechnical Section of NHDOT’s Bureau of Materials and Research.

For more information, visit www.fhwa.dot.gov/innovation/everydaycounts/edc_5/uas.cfm or www.fhwa.dot.gov/innovation/everydaycounts/edc_5/geotech_methods.cfm or contact James Gray at 703–509–3464 or James.Gray@dot.gov, or Ben Rivers at 678–613–2807 or Benjamin.Rivers@dot.gov.