3

Current Practices

The Federal Aviation Administration’s (FAA’s) current approach to safety risk management for unmanned aircraft systems (UAS) is to build on existing concepts of safety analysis and certification for conventional aircraft. This involves “an integrated collection of processes, policies, procedures, and programs used to assess, define, and manage safety risk in the provision of [air traffic control] ATC and navigational services” (FAA, 2017a). This approach has served the FAA and air traffic safety well in the past, but it is possibly outdated with respect to the new challenges and large volumes of operations that commercial UAS present.

FAA uses scalable, multitiered safety targets for different categories of aircraft. The expectation is that airworthiness safety targets may not be the same for all categories of UAS. For commercial transport aircraft, the default “system design” safety target is currently estimated to be one catastrophic event in 1 billion flight hours, but this safety target is less rigorous (e.g, one catastrophic event in 1 million flight hours) for some general aviation aircraft. Within the UAS industry, for example, system design safety targets are different for small recreational UAS and highly customized platforms for in-theater military missions.

The one-in-a-billion (also written 1E-09) safety target for commercial aircraft is based on historical data for which the empirical (or assumed) rate of catastrophic accident per flight is approximately one in a million (1E-06), of which 10 percent of the failures are due to system deficiencies and assuming 100 failure conditions on the aircraft. The product of these yields a safety target of 1E-09. This calculation presumes failures due to “systems deficiencies” and not to other causes (e.g., operations, human errors, weather). U.S. commercial airlines have demonstrated the ability to meet this stringent standard; they have experienced only one fatality since 2009. The presumption of 100 failure conditions, however, is not easily validated with respect to new systems such as UAS. Each of these numbers is a supposition. (The safety target is not the same as the International Civil Aviation Organization [ICAO] target level of safety for vertical separation minima of 5 fatal accidents per billion flight hours (5E-09).¹)

The contributing factors to the concept of the present risk assessment include the following: vehicle design or systems, operational risk, area of operational airspace, the separation strategy, and human versus automation. These, in turn, depend on the established factors of airworthiness, pilot, maintenance, operation, and airspace. Few if any of these risk factors appear to be based on empirical data.

The present approach is to develop a qualitative (ordinal) ranking of probability and consequence for particular categories of UAS operation, and to interpret these rankings of probability and consequence in a so-called risk

___________________

¹ See ICAO Document 9574, para. 1.1.10, http://code7700.com/pdfs/icao_doc_9574.pdf.

matrix or “heat diagram” (see Figure 3.1). Such ordinal rankings are subject to a variety of logical inconsistencies, and good risk practice seeks to avoid them.²

Since small-scale UAS differ considerably from larger-scaled manned aircraft in terms of consequences, the empirical history of their performance should inform a risk analysis.

CLASH OF CULTURES

Safety has been ingrained in the aviation culture from its earliest days. At its birth, aviation was an innovative industry that continually improved the performance of aircraft; to be successful, the industry also needed to enhance the safety of aircraft, crews, processes, and procedures. In its early years, manned aviation had a much higher tolerance for risk than now. Even so, aviation innovators were careful to avoid unnecessary risks. For example, the Wright brothers never flew together. As aircraft grew larger and carried more passengers and as the skies grew more crowded, the tolerance for risk diminished. Government regulations emerged to reduce safety risks, often in response to accidents and incidents. Today’s aviation culture is inseparable from a culture of safety. This relationship is codified in FAA policy; one of the principles of a safety management system is the notion of promoting a culture of safety.³

___________________

² L.A. Cox, 2008, What’s wrong with risk matrices? Risk Analysis 28(2):497-512.

³ For additional information, see the section on FAA Safety Management Policy in Chapter 4.

In the context of aviation, safety is defined as a state where the possibility of harm to people or property is reduced to and maintained at or below an acceptable level of risk.⁴ This safety culture has resulted in a system that provides a commercial transportation capability that has the lowest safety risk of any mode of motorized transportation. While accidents involving large commercial aircraft do occasionally happen, the rate of occurrence is so low that safety experts no longer focus on corrective actions associated with accidents and incidents. They focus instead on proactive safety initiatives based on analysis of precursors of potential accidents. The goal over the next decade is to transition to prognostic safety analysis.⁵

As a result of this safety culture, the aviation community tends to take a conservative stance to new technology and employs an evolutionary approach to change. New technologies are carefully examined to assure that they can meet aviation’s stringent safety requirements. The emergence of unmanned aircraft, especially small unmanned aircraft less than 55 pounds, has tested this culture.

For the most part, the development of small unmanned aircraft is not being driven by the traditional aviation community, but by new participants that have evolved from the electronics and information technology (EIT) culture. Large EIT companies, including Google, Amazon, Intel, AT&T, Facebook, and Verizon, have major initiatives in the unmanned aircraft arena. They are joined by hundreds of new entrants with a Silicon Valley start-up culture. This EIT culture is driven by innovation and is contrasted with the aviation’s safety culture (see Figure 3.2).

While no company would advocate creating an unsafe product, the EIT corporate approach to safety is different. The need to develop innovative products and bring them to market quickly tends to drive strategic decision making and corporate priorities. There is a different threshold and willingness to assume some levels of risk associated with experimenting with new technology, because problems can typically be corrected with software updates after systems are introduced into the market. In other words, innovative EIT companies try new concepts, fail fast, learn, and move on. In contrast, the traditional aviation industry has evolved into one that strives to ensure that safety-critical systems never fail. EIT companies are motivated by competitive market pressures—to take risk in pursuit of potential rewards—and, in some cases, corporate survival. This approach to safety is acceptable for the vast majority of EIT systems, which do not present a direct risk of injury or loss of life if the product does not work as intended. Clearly this is not the case with manned aviation.

___________________

⁴ ICAO, 2013, Safety Management Manual (SMM), Document 9859.

⁵ FAA, 2016, Fact Sheet—Commercial Aviation Safety Team, April 12.

TABLE 3.1 The Contrast Between Small Unmanned Aircraft and Large Manned Aircraft

It is the opinion of this committee that the aviation industry and the public would benefit from an appropriate merging of the diverse safety cultures of the manned aviation industry and the EIT and drone industries. Policy makers have acknowledged that there is a need to be responsive to technology innovation while ensuring continued safety. As FAA Administrator Michael Huerta stated, “No doubt, industry is moving at the speed of imagination. At the FAA, we can’t afford to move at the old speed of government. We have to be willing to innovate the way we do our work, and we are.”⁶

FAA safety inspectors and others involved with certifying aviation technologies and approving operations tend to make decisions based on analysis of years or decades of data and often are focused only on granting permission for a variant on current technologies or practices. However, unmanned aviation safety regulators are often faced with making decisions about technologies where there is little precedent or direct experience (e.g., electric propulsion, fully automated flight control, multirotors, sense and avoid, and network-based communications) and a dearth of data.

Table 3.1 illustrates the wide disparity between small UAS and large manned aircraft. The disparity between other classes of manned and unmanned aircraft, such as large UAS and small general aviation aircraft, are not as stark as those shown in the table. In fact, the cost of entry of a large UAS could exceed the cost of entry for a small general aviation aircraft. Nonetheless, some differences persist regardless of size: the risk assessments for all unmanned aircraft are focused on third parties (since no people are on the aircraft), and the risk assessments of manned aircraft are focused on first parties (i.e., crew and passengers) because the risk they face is very much greater than the risk faced by people on the ground.

Proper identification and classification of the very different safety cases involved in UAS operations and developing, as an industry, the ability to focus on specific use cases and to get their safety analyses is a prerequisite for embarking on further specific use cases. Later in this chapter, an example of the use of small UAS to monitor sea-ice conditions for climatological studies in remote regions of the Arctic is described, where a detailed safety assessment was effectively disregarded. This small UAS use case illustrates how development of priorities for use cases has been, and is likely to remain, difficult. Proper identification and prioritization of use cases should lead to fewer simultaneous UAS standards efforts and standard developing bodies; such focusing of industry resources will inevitably be resisted by portions of the UAS industry.

While the elimination of aircraft accidents and serious incidents remains the goal, it is recognized that the aviation system cannot be completely free of hazards and associated risks. The only absolutely safe aircraft is one that is out of service. Aviation cannot be guaranteed to be free of errors and their consequences. Indeed, the FAA itself has cautioned against driving safety targets to be overly rigorous, as this in fact can reduce the overall level of safety. As Figure 3.3 points out, too much rigor in the targeted level of safety prevents safety improvements from making their way into aviation.

As the FAA is assessing risk for UAS, it is important that an appropriate risk culture be established for this form of aviation. As with small manned aircraft, the collision of a UAS with a manned aircraft poses a threat to human life, with the gravest consequences potentially arising from the collision of a UAS with a large commercial transport. Many UAS safety processes and technology development programs are dedicated to preventing such accidents. Nevertheless, because unmanned aircraft have no humans on board, there is an enormous risk reduction embedded in this form of aviation. A UAS accident does not necessarily mean that a human will be hurt or

___________________

⁶ Michael Huerta, FAA, 2016, “New Horizons,” speech to the Aircraft Electronics Association, Orlando, Fla., April 27.

killed. Furthermore, the public already accepts a background level of risk that is extraordinarily low. The public also accepts the higher level of risk that the crew and passengers of general aviation aircraft currently face, likely because the vast majority of the public does not fly in general aviation aircraft and has no intention of doing so. The public also accepts that medical evacuation helicopters face a risk that is higher still. The level of acceptable de minimis risks varies widely for other societal activities such as traveling by car or motorcycle, swimming in the ocean, or walking across the street. Understanding the level of de minimis risk that the public is likely to accept for small UAS operations, in the context of levels of de minimis risk for other societal activities, would be useful in establishing safety standards for small UAS operations.

UAS provide an excellent opportunity for the development of technologies that can be tested and flown in service that can improve safety for manned aviation at a reduced cost. However, an overly conservative risk culture that overestimates the severity and likelihood of UAS risk can be a significant barrier to introduction and development of these technologies. Specifically, the following behaviors would impede the process of establishing safety regulations for UAS:

• Transposition and assumption of the burden of safety. The burden for safe operations rests on the operator. When the FAA takes on the primary burden for safety, the fear of making a mistake can drive an overly conservative risk culture. Additionally, there is the potential that overregulation actually reduces the competence of regulated organizations.
• Risk avoidance. As stated elsewhere in this report, operation of UAS has many advantages and may improve the quality of life for people around the world. Avoiding risk entirely by setting the safety target too high creates imbalanced risk decisions and can degrade overall safety and quality of life.
• Overanalysis and overreliance on data. When considering the adoption of new technologies, a chicken-and-egg situation can occur where the regulator demands data but simultaneously closes opportunities to collect these data in flight. In addition, in some cases, the expertise and size of the workforce is insufficient to regulate new technologies using traditional approaches.

• Status quo thinking. Sometimes, the regulator will seek to maintain the status quo without acknowledging the technology shifts that are occurring. This is particularly appealing as a safe career choice, given the outstanding safety record for manned aviation. Maintaining the status quo is also attractive if implementing proposed changes would incur substantial costs. However, failure to adapt and make risk decisions based on new technology can have the deleterious effect of keeping helpful technologies out of aviation, missing opportunities to further improve safety and (in some cases) the potential to reduce recurring costs.

Finding: “Fear of making a mistake” drives a risk culture at the FAA that is too often overly conservative, particularly with regard to UAS technologies, which do not pose a direct threat to human life in the same way as technologies used in manned aircraft. This overly conservative attitude can take many forms:

• FAA risk avoidance behavior is often rewarded even when it is excessively risk averse, and rewarded behavior is repeated behavior. Balanced risk decisions are too often discounted: Why risk my career?
• Multiple FAA presenters to the committee stated something to the effect of “we have to protect society” or “society expects the FAA to protect them.” Such a “protect” mentality can result in overconservatism if, for example, it holds UAS technologies and operations to the same standards historically applied to technologies for and operations by manned aircraft.
• Better measures for assessing UAS risk could be considered: Can we make UAS “as safe as other background risks that people experience daily”? And how can the concept of de minimis risk inform the process of assessing acceptable levels of risk posed by UAS? For example, the FAA does not ground airplanes because birds fly in the airspace, although birds can and do bring down aircraft.

The objective is to keep risks under an appropriate level of control, so that they are managed in a manner that maintains the appropriate balance between value and safety. It is important to note that the acceptability of safety performance is often influenced by society norms and culture.⁷ Accordingly, it is appropriate to objectively evaluate the level of risk the public is willing to accept with respect to UAS (e.g., flying over people, critical infrastructure, etc.) and to consider the results of this evaluation when establishing risk-based standards and regulations. As part of this evaluation, assisting the public with understanding risks that are avoided by UAS operations is important. As stated earlier, this approach has worked well for the FAA in achieving a very high level of safety for air transportation.

Reviewing approaches used by other nations is also informative. For example, Sweden requires equipping all UAS with an emergency shutdown capability. In France, under certain circumstances, some UAS are required to have an automatic system to prevent them from going beyond a specified distance from the operator (Law Library of Congress, 2016).

As highlighted above, the FAA administrator recognized several years ago the need for government to move faster in addressing the burgeoning drone industry. But how should that recognition be put into action? The first step is to listen to and collaborate with industry. The FAA is doing that through many venues, including the National Academies of Sciences, Engineering, and Medicine and this committee. The FAA has established the Drone Advisory Committee and numerous aviation rulemaking committees through which they work closely with their stakeholders. The FAA also works with organizations such as RTCA to develop minimum performance standards that serve as a means of compliance with FAA regulations. The use of performance standards rather than prescriptive design standards encourages industry to develop innovative designs and solutions that comply with the standards. This approach has never been so important and pertinent as it is now when applied to new entrants into the airspace such as small drones.

There are opportunities to further incorporate these innovative and collaborative approaches into the FAA’s internal culture. Doing so could lead to an environment in which FAA personnel charged with any part of the regulatory process are encouraged to accept reasonable risks rather than avoid action as a way to avoid accountability and negative impacts on their careers. This could also lead to the creation of a proactive safety culture that

___________________

⁷ ICAO, 2013, Safety Management Manual (SMM), Document 9859.

looks for how to get to “yes” without compromising safety, rather than one that dwells on what might go wrong. A system that rewards finding ways to enable new operations and penalizes inaction is the best way to jump-start the cultural change needed to “move at the speed of imagination.” Further, in a proactive safety culture, responsibility, authority, and accountability are clearly articulated for each member of the safety organization, and no one person can derail an initiative simply by not saying yes; this is in contrast to a culture of management by committee or internal boards or panels.

PROCESSES

The current FAA process for considering and approving routine UAS operations (see Chapter 2) continues to stifle needed industry investment in developing technical and operational risk mitigations. The lack of empirical data continues to be the driver for the agency’s subjective approach to approvals. Each request requires, even for “one-off” operations, an extremely labor-intensive, detailed description of an operational plan and system description. Commercial operations and public UAS operational approvals differ somewhat, but both are subject to significant scrutiny prior to approval with scant guidelines for how to show compliance with rules and regulations. The current Certificate of Authorization (COA) application, required for public aircraft operations, requires detailed descriptions of more than 15 separate items (see Table 3.2). Although the FAA’s “Accountability Framework” clearly states that the responsibility for having safe airborne systems is with operators and manufacturers, with the FAA providing oversight, there is a strong culture within many parts of the FAA that it is the FAA that is responsible for airborne platform safety. This culture has led to highly prescriptive FAA guidance for many

TABLE 3.2 Details of Certificate of Authorization Information Requirements

NOTE: ATC, air traffic control; DOD, Department of Defense; FAA, Federal Aviation Administration; GPS, Global Positioning System.

SOURCE: Adapted from FAA (2008).

aspects of airworthiness certifications. While the FAA is attempting to “streamline” certification in a number of certification domains, the streamlining initiatives face ongoing internal debate and resistance within the agency as well as expectations for harmonization with international airworthiness authorities (e.g., ICAO Cir. 328, ICAO A39-WP/116).⁸

If approvals are granted, they are valid only for specific operations over a finite period, are subject to continued FAA scrutiny, require data sharing with the agency, and do not apply to commercial operations. Approvals are granted only after internal FAA discussion and risk assessment, because there are few if any specific performance standards available to serve as a means of compliance with rules and regulations. The only exception to the above process is for emergency approval issuance in the case of significant threat to life or property (natural disasters or other emergency applications).

Civil or commercial operational approval requests can be even more daunting and uncertain as to the probability for success. In addition to the above, civil airworthiness requirements are added to the process. Experimental, Restricted Category, or Special Airworthiness certification is normally required. The approval process involves the action of a formal safety risk panel if the national airspace is impacted and pertains only to specific operations. Routine “file and fly” operations of commercial UAS are still essentially prohibited.

The decision-making process within the FAA for a Safety Risk Management Document (SRMD) requires sign-off by Senior Executive Service (SES) personnel in multiple FAA organizations. In a recent case, the SRMD for UAS Detect and Avoid (DAA) Safety Assessment acceptance, submitted to the FAA in May 2017, required SES personnel from 12 organizations to sign off on the document before it could be accepted. In this case, not a single FAA person has signed off on the document as of February 2018. This is true, despite the fact that (1) the FAA initiated the creation of the DAA Minimum Operational Performance Standards (MOPS); (2) the FAA was integrally involved in the creation of the MOPS; and (3) the estimated cost of the development of the MOPS is approximately $250 million (including NASA verification and validation work, safety and safety assessment, and all the resources brought to the RTCA Special Committee-228 to complete the MOPS).

Recommendation: The FAA should meet requests for certifications or operations approvals with an initial response of “How can we approve this?” Where the FAA employs internal boards of executives throughout the agency to provide input on decisions, final responsibility and authority and accountability for the decision should rest with the executive overseeing such boards. A time limit should be placed on responses from each member of the board, and any “No” vote should be accompanied with a clearly articulated rationale and suggestion for how that “No” vote could be made a “Yes.”

At present, unmanned aircraft designed to fly beyond visual line of sight for commercial purposes require a formal airworthiness certification. To date, only two viable paths exist to accomplish this: either Restricted Category or Special Class Type Certification. Both are difficult to obtain and have restrictions and limitations associated with them, and neither guarantees access to airspace. Special Class Type Certification under CFR 14 Part 21.17(b) has yet to be granted to any UAS, even after more than 2 years of effort by the proponents. The man-hour expenditure associated with obtaining the currently required certification and operational approvals generally well exceeds the value of the majority of commercial business opportunities. The expense and uncertainty associated with meeting the vague “risk-based” requirements imposed by the FAA make it difficult if not impossible to compete with certified manned aircraft serving the same mission.

A key product sold by commercial UAS service providers and manufacturers is data. UAS are operated under the assumption that they would offer an efficient and safer method to collect information for a customer. The systems are designed and operated to carry sophisticated payloads capable of capturing a wide variety of data and imagery in missions considered too “dull, dirty, or dangerous” for manned aircraft. Unfortunately, the onerous requirements to obtain approval to conduct such missions have resulted in many UAS service providers transfer-

___________________

⁸ ICAO, 2011, Cir. 328, “Unmanned Aircraft Systems (UAS)”; ICAO, 2016, A39-WP/116, Working Paper, “The Need for Standards in Support of Harmonized UAS Operations.” Presentation, https://www.faa.gov/uas/resources/event_archive/2017_uas_symposium/media/Breakout_3A_Global_Leadership.pdf, is a good example of the efforts for international harmonization.

ring their payloads to manned aircraft to meet customer requirements. The ScanEagle UAS (a Group 2 UAS with over 1 million logged flight hours) (see Figure 3.4) is one example of a commercial business that is structured around data collection utilizing a combination of manned and unmanned aircraft when it would be safer and more cost-effective to conduct operations solely with unmanned aircraft.

In all likelihood, requirements for additional equipage—to be compliant with detect-and-avoid and command- and-control standards to operate routinely in the national airspace—will add to this burden. The addition of air-to-air radar, additional command-and-control capability, and increased aircraft performance to meet collision-avoidance requirements will likely force redesign of existing UAS in order to accommodate the additional size, weight, and power necessary to comply. All of these issues add to the challenge of closing a business case for UAS in the near term.

Commercial/civil operations can be categorized in two market segments. The most widely enabled are those conducted by small UAS (vehicles weighing less than 55 pounds) operating at altitudes below 400 feet above the ground and within visual line of sight of the pilot. These operations serve rather modest markets, such as real estate surveying, news media, some first responder activity, and localized precision agriculture. With FAA’s current “risk-based” approach to operational approval, these types of operations are accommodated by compliance with the fairly easy requirements of 14 CFR Part 107.

The larger, more robust market opportunities exist in operations characterized by higher altitude, beyond visual line of sight, long-duration, and linear surveillance applications. The FAA considers these operations much riskier because of the size and performance of the vehicles needed to carry them out and because aircraft conducting these operations typically will share airspace with passenger aircraft. To the FAA’s credit, emergency operation of these larger vehicles has been approved in response to some recent natural disasters. However, routine use for such applications as long-distance power line and infrastructure inspection, large-tract agriculture, wildfire monitoring, oil and gas pipeline inspection, severe weather monitoring, search and rescue, and law enforcement has been highly restricted.

Although there may be additional risk associated with operations of larger, higher-performance vehicles, the operations most attractive to industry are normally conducted over very remote locations where the risk of encounters with other aircraft or people not associated with operation is sufficiently low. Unfortunately, the current anecdotal approach to assessing the risk and the associated uncertainties of gaining operational approval make it difficult to establish a sustainable business model. Accordingly, industry continues to be wary of making additional aggressive investments in the technology.

The FAA has underutilized the test sites, pathfinder programs, and the upcoming UAS Integration Pilot Program by not defining and collecting data that could inform risk assessments. There are numerous examples of how UAS could be used to deliver emergency services to people in need:

• Delivering life preservers to swimmers in lakes or the ocean (extended visual range),
• Delivering automated external defibrillators to distressed persons in state parks,
• Searching for lost hikers in national forests, and
• Monitoring ice and tracking whales in remote marine environments off the coast of Alaska (beyond line of sight).

Finding: The safety of the National Airspace System has been achieved in large part as a result of the FAA’s risk decision process, which has been characterized by a culture with a near-zero tolerance for risk. This culture, however, has too often resulted in overconservatism in the SRM process as it has been applied to UAS technologies and systems. The SRM process is particularly vulnerable to overconservatism due to its subjective nature. In particular,

• An overly conservative culture prevents safety-beneficial operations from entering the airspace. The focus is on what might go wrong. More dialogue on potential benefits is needed to develop a holistic risk picture that addresses the question, What is the net risk/benefit?
• Paralysis by analysis, where more data are requested in light of uncertainty about new technology, but flight experience cannot be gained to generate these data due to overconservatism.
• The status quo is seen as safe. There is too little recognition that new technologies brought into the airspace by UAS could improve the safety of manned aircraft operations, or may mitigate if not eliminate some nonaviation risks.

PROCESSES THAT DO NOT WORK

During its deliberations, the committee heard of numerous examples where proposals to use UAS in ways that were only slightly changed from previous practices met lengthy delays and were ultimately rejected for reasons that the proposers could not understand. Consider, for example, the Marginal Ice Zone Ocean and Ice Observations and Processes Experiment (MIZOPEX). This experiment was a $3.5 million project funded by NASA with the goal of helping to address information gaps in measurements of basic parameters, such as sea surface temperature, and a range of sea-ice characteristics, through a targeted, intensive observation field campaign that tested and exploited unique capabilities of multiple classes of UAS. To help achieve this goal, the experiment as designed included the use of a UAS weighing just 1.5 pounds and flying at a maximum height of only 50 feet over water in a very low traffic area north of Alaska. After a yearlong review, the COA to include this UAS in the experiment was denied.

Appendix C provides a detailed description of the experiment and the long sequence of events that prevented the MIZOPEX field campaign from including what seems to have been very low risk flights by a small DataHawk UAS (see Figure 3.5). One of the participants in the project, J.A. Maslanik, summarized lessons learned during the MIZOPEX project as follows (Maslanik, 2016):

This example, unfortunately, is not unique and is one reason why news of many of the most innovative uses of UAS often comes from non-U.S. locations. In Chapter 4, the committee offers recommendations on how to improve this situation.

REFERENCES

DOE (Department of Energy). 2013. Polar Research with Unmanned Aircraft and Tethered Balloons: A Report from the Planning and Operational Meeting on Polar Atmospheric Measurements Related to the U.S. Department of Energy ARM Program Using Small Unmanned Aircraft Systems and Tethered Balloons, September. https://www.arm.gov/publications/tech_reports/doe-sc-arm-tr-135.pdf.

FAA (Federal Aviation Administration). 2008. Sample COA Application, V 1.1. https://www.faa.gov/about/office_org/headquarters_offices/ato/service_units/systemops/aaim/organizations/uas/media/COA%20Sample%20Application%20v%201-1.pdf.

FAA. 2017a. Safety Management System Manual. https://www.faa.gov/air_traffic/publications/media/ATO-SMS-Manual.pdf.

FAA. 2017b. Presentation by B. Crozier, FAA, to the National Academies Committee on Assessing the Risks of UAS Integration, September 26.

Law Library of Congress. 2016. Regulation of Drones: Australia, Canada, China, France, Germany, Israel, Japan, New Zealand, Poland, South Africa, Sweden, Ukraine, United Kingdom, European Union. https://www.loc.gov/law/help/regulation-ofdrones/regulation-of-drones.pdf.

Maslanik, J.A. 2016. Investigations of Spatial and Temporal Variability of Ocean and Ice Conditions in and Near the Marginal Ice Zone: The “Marginal Ice Zone Observations and Processes Experiment” (MIZOPEX) Final Campaign Summary, DOE/SC-ARM-15-046. Ed. Robert Stafford, DOE ARM Climate Research Facility. https://www.arm.gov/research/campaigns/osc2013mizopex.