As a frenzy of activity continues to circulate around the use of unmanned aerial systems (UAS), insurance companies are researching the viability of their use, including looking at the opportunity to minimize injuries to personnel climbing roofs to review damage and, in the event of widespread disaster, being able to quickly survey damage and prioritize claims resolution operations.
Many carriers have actively initiated programs, going beyond research into the areas of regulation and testing. Early adopters are focusing on different types of UAS that may be ideal for resolving claims in the field and visualizing risk. As the industry moves toward the use of UAS, it is important to remember that it is about more than just using the unmanned aerial vehicle; it is also about the systems, processes, and technologies supporting its use.
As regulations try to catch up with the technology, it is not a matter of "if" your company will be impacted by the use of drones; it is simply a matter of "when." As such, understanding how the collected UAS data integrates into current workflows is imperative and has been a focus of the recently formed Property Drone Consortium (PDC).
The consortium is committed to helping address many factors surrounding data collection, processing, and delivery, along with safe and economical capture. The PDC represents a collaboration among insurance carriers, construction industry leaders, and supporting enterprises that have agreed to work together to promote research, development, and the establishment of regulations for the use of UAS technology across the insurance and construction industries.
For the insurance industry, many have agreed that collaboration and shared research is the best allocation of resources in terms of internal skill set, regulatory insight, financials, and the overall technological expertise with aerial platforms, sensors, data capture, data management, and workflow integration, most of the aforementioned being outside of a carrier’s core competencies. The consortium continues to focus on the current environment but also, and more importantly, on what carriers and other businesses will need once drones are legal for expanded commercial use.
Current Regulatory Environment
With new exemptions, research, and information being released daily, the larger picture and potential long-term use of this technology changes frequently. In recent months, both trade and consumer media have provided an enormous amount of coverage revolving around developments in the UAS industry. The flurry of activity by the Federal Aviation Administration (FAA) with its charter to integrate drones into the U.S. national airspace (NAS) for commercial purposes has been highly publicized. In February 2015, the FAA proposed long-term rules for the professional use of unmanned aerial vehicles (UAV), also referred to as drones and microdrones. Since then, the industry has anxiously awaited a final set of rules, which the FAA recently submitted to the Office of Information and Regulatory Affairs (OIRA) of the Office of Management and Budget (OMB) and expected would be published by the end of June 2016.
As the FAA has waited for the final rules to be published, it has made some strides in streamlining the current approval processes for research and development. It has established six UAS test centers that can issue certificates of authorization (COA) for research and development and has streamlined the process for seeking and receiving an FAA Section 333 exemption that grants permission to fly. Current regulations allow for the use of UAS for research or commercial purposes under strict provisions, e.g., distance from certain airspace, distance from populated areas, above ground level restrictions, distance from nonparticipants, and pilot and observer qualification requirements. Under these new processes, over 5,000 Section 333 exemptions have been granted, although it appears that the FAA is inundated with requests and approval times are taking as long as six months.
The FAA also has significantly simplified the registration process for UAS platforms with its online registration system and in recent months has announced task force recommendations for the use of platforms in the microdrones category (less than 4.4 lbs.). It also has simplified requirements for students using drones as part of their studies and established a long-term Drone Advisory Committee consisting of industry stakeholders tasked with examining key issues relative to the safe integration of UAS into the NAS.
Despite the improvements the FAA has made in the current process, it can still be time-consuming and costly, and the provisions for commercial use or for research and development can be very restrictive. It has become obvious that companies looking to establish a market presence or incorporating drone usage into their business will need to pay a great deal of attention to regulatory issues and ongoing developments.
However, there is far more to flying a UAS than just gaining regulatory approval. A significant number of factors need to be addressed, including who will fly the vehicles and which platforms to use. A critical factor is determining the types of sensors and cameras to use and identifying the data they will be able to collect, process, and deliver. Overall, the importance of the data is monumental, and unless captured, processed, and analyzed correctly, it will lead to a huge waste of resources. Data management and the integration of derived information into workflows that maximize ROI is a prime objective for insurance carriers.
Success Factors: Platform Selection
It is only logical that any discussion about UAS start with consideration of the platform itself. The size and capability of the platform can greatly simplify the operator skills required, but at what cost? Each added stability and control automation subsystem adds to weight, battery depletion (or other fuel consumption), and complexity. The question should be asked: Are we going to capture single structures one at a time, or is there a need to capture neighborhoods?
If individual structure inspection is required, then a multirotor device is the optimal platform choice. This type of platform operates closer to a target at lower airspeeds, and it is less prone to image blur because the drone can hover. The operator can control distance easily in order to capture the best view. Battery life is a limiting factor, as rotors alone lift the weight of the device. Ten to 20 minutes per flight is reasonable, and the user should consider having extra batteries available.
If broader or neighborhood area coverage is required for assessment, post-disaster inspection, or other uses, a winged drone may be the better choice. This type of platform can stay aloft longer and cover a broader area. However, winged drones need airflow for lift, and it is likely that a fix-winged system would require beyond line of sight operation, which presents its own set of regulatory and technological hurdles. For example, UAV transmitter range may become a larger concern when trying to cover a large area. Flight with, against, or across the wind can cause the device to vary greatly in effective groundspeed. This will drive image quality down and hamper the ability of the operator to control the device. Winged drones, in general, will be a single engine with electric power. As they have a lifting surface, they are often made of a frangible material capable of impact absorption. A strong impact will often destroy the wing, tail, or airframe. Different materials have different impact resistance. This is a design feature, and replacing foam material wings or a fuselage will be part of the operating costs.
Rotary drones come in a multitude of forms. They may be single axial rotor(s) with or without tail rotors. In appearance, this is the look of a traditional manned helicopter. They may be multirotor in design, with many rotors distributed across the airframe to provide lift and control. Configurations of four, six, or eight rotors exist in many frame styles. More rotors may add complexity but can mitigate risk to some degree should a single rotor fail.
Success Factors: Image Capture
Just as the platform vehicle can greatly affect the clarity and efficiency of capture, the sensors and cameras that are used are critical. There are many types of passive and active sensors to mount on a UAV, depending on the data capture requirements. Options include still photography in black and white or color, as well as video. The type of system is a major consideration. The difference between frequently captured still images and video can be dictated by storage capacity and intended use of the imagery. The storage and delivery of imagery is another area requiring strong research.
Video streaming is more likely to ensure usable capture, but there are some risks as to the quality of resolution as compared to a still frame data set. Multiple concurrent captures of streamed and stored onboard data can exist but, again, adds to complexity and weight. In either case, a first-person view (FPV) system provides an added measure of assurance for accurate data capture. FPV does require an added operator, as the pilot in charge must always have an unassisted view of the UAV due to current FAA requirements on visual line of sight to the vehicle. It is imperative that the user reviews the data captured before leaving the site, extracting it from onboard storage and viewing it immediately after landing.
Success Factors: Sensor Selection
As noted above, there are many different types of active and passive sensors to choose from in addition to traditional cameras. These sensors detect light in various parts of the imaging spectrum, from ultraviolet all the way to thermal infrared. They have varying degrees of spatial and spectral resolution for use in different applications. For example, one common UAS spectral application is the detection of vegetation health. Other potential applications include material identification, permeable surface detection, and thermal leakage analysis.
The PDC is actively researching what needs to be seen within the imaging spectrum. Inspectors will inherently understand what they can see from different vantage points looking at a roof with a handheld camera or the naked eye. For example, they know what they can see in shingle deformation when the sun is high with shadowing. Would it be worth learning how to interpret images captured beyond the spectrum of the human visual system? If we added the capabilities of digital sensing, we might be able to see mold accentuated to levels not usually seen through the red-green-blue portion of the spectrum. It may be possible to characterize roof materials, potentially differentiating polymer versus oxidized roofing materials. This could lead to a process for generating reports and showing degree of aging across a roof. All of these might be possible if we consider technology and applications showing the importance of looking well beyond simply flying.
The selection of the right sensor produces powerful capabilities. If we use a georegistered sensor, we also capture data about its location and the area it is recording. Data captured from a drone will likely be merged with other imagery derived from manned missions and possibly satellite, offering more than one temporal look at the target. We can see how the roof in question has aged and how it compares to others in the region. We also will be able to see if the owners have made changes over time or after storm events. The sensors’ georegistration means that a spatial archive captures location and addresses enabling users to acquire and analyze a wealth of information quickly and easily.
Success Factors: Data Management
Once captured, how does an insurance carrier use the data to make business decisions? Is the data strictly for visualization, or is it authoritative? Where is all the data stored, and how is it retrieved? The amount of storage needed for this type of collection is tremendous. Choosing functional platforms is just the beginning of a much larger issue of utilizing accurately captured data in the workflow for efficient claims review and risk visualization.
The systems need to be able to capture intelligent images that integrate into a strong, predetermined process for maximum ROI of the data and then be able to store and retrieve the same data. With the correct systems, this can be automated, flowing imagery and data into current software viewing tools that can measure anything visible while utilizing a number of other analytic and GIS tools to append layers, review parcel data, perform change analysis, and create 3D models. Visualization means that a user can see and interpret the content. By making it authoritative and tracked to the data, processes can be added to include assurance of privacy and allow the data to become a legal record for future use.
Collaboration and the Property Drone Consortium
The primary key to risk mitigation is trustworthy data coupled with credible research. While UAS have been in use for a number of years in the military, little data exists in the commercial sector to address the possible risks with the integration of UAS into the national airspace. Therein lies the dilemma for the FAA and a number of insurance carriers as they contemplate the integration and the use of UAS.
While watching the progress of the FAA, carriers are preparing for the potential use of UAS. Each carrier has priorities and limitations, but there are certain key fundamentals to contemplate when considering UAS for commercial use. It is about a great deal more than just getting FAA approvals and flying the platform. As noted earlier, there currently are over 5,000 Section 333 exemptions that have been granted as well as a number of COAs approved via UAS test centers, municipalities, and academic institutions.
The time is rapidly approaching in which a world with commercial drone use will be commonplace. The competencies required for developing applications of this technology are many, even within a single industry. While it is important to develop an understanding on an individual level, collaboration with an organization such as the PDC will help address many, if not all, considerations and success factors.
The author would like to thank the following for their contributions to this article: Charles Mondello, David Nilosek, and John Monaco.