The How’s and Why’s of Mapping With Drones
Anyone actively involved in the profession of surveying and mapping has probably heard a lot of talk about the use of unmanned aerial systems (UAS) as a way of providing mapping services to clients. In the past couple years since the Federal Aviation Administration (FAA) and U.S. Congress made large areas of the country available to unrestricted use of UAS technology, many people in the geospatial professions have already hired someone to provide these mapping services. Some of us have become FAA-certified remote pilots, flying UAS ourselves and providing mapping services.
However, there are many questions on the use of UAS in geospatial: How and why is it being done? What accuracies can be expected? Who can legally provide these services? What are the rules and restrictions surrounding the use of UAS technology? Let’s take the five W’s and the H one-by-one.
01 What is UAS?
A UAS consists of several hardware and software components:
- The unmanned aerial vehicle (UAV) is also commonly called a DRONE, which is an acronym for dynamic remotely operated navigation equipment.
- Flight planning software allows the pilot to predefine a flight path for the UAV to follow during autonomous flight mode
- The radio controller connects flight plans with the UAV, and provides communication between the remote pilot and the UAV.
- The controller allows the remote pilot to fly the UAV in manual flight mode where each touch provides direction to the UAV.
Camera systems are a very important consideration when using a UAV for mapping applications and are the main reason for the use of UAS – photographs. Choosing the correct camera is a key decision when considering the positional accuracy of the products that are required of a UAS flight. Lastly, post-processing software is required to turn photographs into imagery that can produce accurately scaled image mosaics (orthophotos), three-dimensional planimetric and contour topographic mapping.
02 Why Use UAS Technology?
Why do anything in business? Because it adds value, provides a new way to produce a product, provides a new product the client can appreciate, it’s safer than the other way of doing things and it improves profits. UAS is not the answer to all mapping applications, but when it is one of the answers, it should be considered.
03 Who can Provide the Services?
Anyone can fly a UAV as a hobby, but as soon as UAV products and services are acquired for a fee, the operation becomes a commercial designation. At this point, the FAA and federal law require the UAV pilot to be certified. Certification requires a background check and passing a knowledge test at an FAA-approved facility. The FAA does not require remote pilots to carry insurance for their operations; however, as with any consultant you work with, your UAS consultant should carry insurance too.
Just because your UAS consultant has a remote pilot certificate and is insured does not necessarily mean they are equipped to provide quality mapping services. You will need to know the answers to the following questions:
- Do they have experience?
- Do they have the software to generate all the products that you require?
- Do they understand the accuracy requirements of your project?
- Can they be clear about where your site conditions (areas of trees, water, urban canyons, etc.) may provide accuracy challenges?
- Does the consultant process the imagery themselves, or do they send it to the “cloud” for product generation?
- What checks can they perform that provide confidence in the product accuracy?
- Can they properly say and spell the word “photogrammetry?” (i.e., do they meet local licensing requirements?)
Just as you would probably not hire an inexperienced company claiming to provide “manned aerial photogrammetry” services, you should consider the qualifications of a UAS consultant just as closely.
04 Where can a UAS be Used?
The FAA considers the airspace from the ground to 60,000 feet above ground level (AGL) to be under their control. That’s right… the air that is inside your property line is not yours, it is the federal government’s. A whole book could be written about airspace, but the very short version is that the country is broken into several classes. Class B, C and D airspaces exist near “major” airports. These are titled as a “controlled airspace” and require approval from the FAA and air traffic control to operate UAVs in. Approval can take 30-60 days on average, making operations in these areas difficult within tight schedules. In some cases, approval will not be granted.
Class G airspace is “uncontrolled” and generally open to unrestricted use by UAVs. With a few exceptions, Class E airspace exists at an altitude above the operational limits of a UAV, and Class A airspace is where commercial air traffic takes place above 18,000 feet.
Lastly, the FAA requires that UAVs “cannot be flown higher than 400 feet AGL unless flown within a 400-foot radius of a structure and does not fly higher than 400 feet above the structure’s immediate uppermost point.” There are other restrictions on where UAVs can be operated, but a certified remote pilot should be able to provide guidance as to where. A great majority of the country is available for UAS use.
05 … And When?
Night flying is prohibited. It is possible to get FAA approval to fly a UAV at night, but a restriction waiver is required. Unless you need a thermal infrared imagery product that can only be collected at night, daytime operations are the preferred time to gather photographs.
When manned flight aerial photo missions are needed, it is important to closely monitor sunny skies and solar angle. This is truer for manned flights than unmanned. Manned flights for mapping projects are usually conducted between 1,000 and 5,000 feet AGL. The large format cameras and downstream processing of the imagery work best with proper illumination and shadows. Also, flight operations are very difficult to manage under partly cloudy conditions when the camera is (or could be) higher than the cloud deck. With UAS operations being conducted below 400 feet, partly cloudy to overcast conditions do not pose a problem if there is enough light to properly expose an image. In some cases, overcast conditions are a benefit as there is little chance of the imagery being too overexposed or too “hot.”
Shadows are not as important when UAS imagery is flown because the resolution of the photos is so good, a human or machine operator can much more easily interpret the imagery than they can from high altitude photos. As a guide, UAS imagery is typically produced with image pixels that are 0.5 to 1.5 inches square. Manned aircraft image pixels average between 1.5 to 6 inches square.
Lastly, UAS technology allows for the aerial photographic workflow on project sites that are economically impractical to use a manned aircraft approach. If the end products support the project requirements and provide benefit to the client, UAS technology is a great option.
How does UAS Technology Work?
The answer is, of course, “It depends.” Do you need still photos or a video of a construction site or project area? If so, almost any aerial camera platform will provide these products with very good results.
Do you need contour and volumetric analysis on stockpiles, but don’t care if the mapping data is tied to a coordinate system? If so, many consultants can take photos, upload the imagery to the cloud, and have a third party generate an elevation model from which quantities can be extracted. Bare earth (no vegetation or water) conditions are the easiest areas to map. The resulting point cloud usually represents the ground very well and accurate volumes can be extracted. Of course, if the consultant does not post-process the imagery themselves, it may be difficult to get a report showing the accuracy of their point cloud and resulting volumetric quantities.
You might ask, “Aren’t point clouds the result of light detection and ranging (LiDAR) surveys?” They are, but point clouds can also be generated using photogrammetric technologies. Post-processing software uses a technique where all image pixels from one photo are compared with matching pixels in adjacent photos. When common matching features are found in two or more photos, a horizontal and vertical point is calculated for each image pixel. The points can be expressed in the American Standard Code for Information Interchange file format (ASCII), as well as the binary LASer (LAS) file format that is typically used to distribute LiDAR data. The result is a file generated with x, y, z coordinates.
Do you need 3D contour mapping and orthophoto data that is accurately tied to a specific coordinate system? If so, your UAS survey must use surveyed control points or an airborne GPS survey using an active base station. Alternatively, a UAV that integrates very accurate inertial measuring units (IMU) and global satellite navigation systems (GNSS) allows for the reduction or elimination of physical control points on the ground. The post-processed photos will need to undergo the pixel image “matching” procedure mentioned above, as well as combining measurements at physical control points, or the adoption of GNSS and IMU inputs from the UAV, so the x, y, z point cloud can be accurately positioned to a known coordinate system. All measurements derived from the processed photos and point cloud will then be tied to the desired coordinate system. If your UAS consultant cannot provide surveying services, an additional consultant may be needed to provide the ground control points.
Do you need 3D contours, orthophotos and planimetric (feature) mapping data that is accurately tied to a specific coordinate system? If so, see previously mentioned points No. 2 and 3 and ask yourself, “How will a UAS consultant accurately provide the locations of such things as roads, buildings, hydrology, vegetation and other linear features?” They will need to be digitized somewhere into a CAD file of some sort (AutoCAD, MicroStation). Unlike terrain mapping, there are no practical ways to automate the extraction of lines and points from photos. They need to be manually digitized and placed on a CAD layer. There are two ways to do this. One is to insert an orthophoto into a CAD file and trace all features on a 2D monitor using “heads-up digitizing.” Try it some time. Add a photo to a CAD file and trace everything you see with a 2D mouse. Make sure you snap all the end points and properly code all the objects. Do you need parallel strings on the curbs? Do it. It is not fun.
The second way is to insert all the aerial photos into the same 3D stereo collection (softcopy) software that is used to compile maps from manned flights. Using this approach, you can compile the features directly into a CAD file without the need to look at an orthophoto. Also, consider that while the orthophoto is horizontally accurate, there is no way to digitize accurate 3D line work or point features from them. Only from a stereo environment can breaklines and point elevations be digitized exactly where they are needed to further support the surface that was generated through the “matching” routine. It is possible to compile 2D lines from an orthophoto and then compute the elevations on the line from the elevations found in the point cloud (draping). Exact placement of a line that represents an important breakline feature will be hard to do and might support the 3D terrain very well. However, without the ability to inspect the automatically generated surface or 3D elevated lines computed from that surface in a stereo mode, how can we be sure if the point cloud features are accurate? Just because a piece of software generates a surface does not mean it is perfect. It must be edited for erroneous points that fall near vegetation, water and all other obscuring features.
Tell Me More!
The great thing about UAS mapping is the very high-resolution imagery and the relative simplicity of processing the imagery into surface models and orthophotos. The real weakness in UAS mapping is the difficulty checking the accuracy of the surface model and digitizing the planimetric features.
As with all new technology, it is a good idea to look for creative ways to adopt it. Become familiar with the good, the bad and the ugly. Compare the new technology to the old and learn how to make the most of it moving forward.