Perspectives: Commercial use of drones could be grounded by costsReprints
There is significant tension between the Federal Aviation Administration regulation for commercial drones and any conceivable economic model of a profitable retail drone delivery operation.
Ultimately, full automation may be required to realize that dream, which is a reality at odds with the current regulatory requirement of having a pilot in command of a drone. Efforts to achieve regulatory changes will draw the FAA’s priority of aviation safety into some level of compromise with the financial priorities of commercial retail drone operators — a balance that could create significant risks for retail drone operators if those compromises go too far.
Line of sight
In June 2016, after years of public comment, the FAA released the long-awaited Part 107 commercial drone regulation. The regulation reflects a cautious and preliminary approach to commercial drone operation.
Among other things, the new regulation mandates that operations must occur during the day and that a pilot in command under “visual line of sight” conditions operate the drone, avoiding flying over humans.
The regulation creates a remote pilot in command position but limits operation of a remote drone to one at a time.
Part 107 also creates a procedure for obtaining waivers from the regulations.
However, the provisions specifically state that the VLOS requirements allowing operators to carry property for compensation or hire will not be waived. Furthermore, such carriage of property is limited to operations conducted within a confined area and in compliance with the operating restrictions of Part 107.
Any waivers, therefore, that the FAA may grant to the VLOS requirements under Part 107 would not allow the drone operation to transport property for compensation or hire beyond VLOS. Accordingly, the regulation does not, by itself, provide the necessary environment for retail drone operations, even accounting for the waiver process.
The absence of rules relating to overhead operations was not an oversight by the FAA. In early 2016, the FAA sought input from a variety of drone industry stakeholders “to consider recommendations for a performancebased standard that would allow micro (unmanned aircraft systems) to be operated over people.”
The stakeholders — serving as the Micro Unmanned Aircraft Systems Aviation Rulemaking Committee, or ARC — included several companies and associations involved with the commercial use of drones in several different contexts.
The task force developed and proposed a framework for operations over human populations, which was published on April 1, 2016. This included four “performancebased standards for the classification of UAS operated over people,” each of which would permit an increased degree of operation over people subject to certain operating requirements.
Material to each category was an assessment of risk in light of probable failure modes. The “acceptable probability” of risk was measured in relation to the chances of an “(Abbreviated Injury Scale) level 3” injury — significant injury or death — caused by the impact of a falling drone. A drone operator would have to certify that its operations would not exceed a specified risk threshold. Categories 3 and 4 contemplated a maximum 30% chance of causing an AIS level 3 injury. In other words, significant injury was acknowledged as a calculated cost of operating drones over human beings.
These performance-based standards and their corresponding risk levels were not incorporated into Part 107 and implicit in the omission is the FAA’s concern for safety. Indeed, Part 107 flatly precludes operation over human population. While the ARC standards were not incorporated, they do illustrate the type of analysis that will be required once commercial operators seek approval to operate over human populations. The necessity of calculating an “acceptable” risk of “significant injury or death” is an inescapable aspect of the analysis.
Openly calculating the risk of significant harm to people should alarm anyone, especially those associated with aviation safety or corporate risk management. The standard for punitive damages in most jurisdictions requires a showing of knowing disregard for the safety of others. In attempting to meet that standard, it would be very helpful to plaintiffs lawyers to have risk assessment data and documentation that shows the defendant took a formulaic approach to calculating risk to humans and determined what it thought to be an acceptable range.
It should also be noted that data utilized for waiver applications would also be discoverable in any litigation relating to a drone accident. Risk assessments and investigations of any accidents related to testing operations will draw attention of those bringing such claims. Managing that data and analyzing it properly — with appropriate consideration given to the risks to the public — will be critical to minimizing exposure in future litigation.
The problem of economics
Nevertheless, the pressure to change the current regulatory scheme is largely economic.
Consider for a moment the costs associated with the delivery of a six-pack of beer by drone. Let’s say that the item costs the customer $15. Added to that cost might be some portion of an annual subscription fee for airborne delivery, such as Amazon’s annual $99 for its Prime services, raising the revenue allocated to that single delivery to something like $20. Now consider the expenses an accountant would likely allocate to this single delivery flight, including the wholesale cost of the beer as well as some allocation of research and development, lobbying, administrative — for example, costs related to obtaining waivers — insurance, materials, inventory and other expenses that appropriately factor into net profit.
In order to make the process profitable, one has to substantially increase the volume of deliveries. Increasing a human pilot’s monitoring burden to, say, dozens of flights would not only dramatically increase the risk of pilot error but it also would fail to remove the single most significant cost involved in the process: the pilot.
Seen in this light, it is not hard to understand the correlation between fully automated drone operations and profitability.
An automated system that can monitor hundreds of deliveries at a time would over time prove to be the optimal model for a retail drone enterprise, even if there were some human oversight involved by a few employees. However, by doing so, it removes what the FAA would consider to be the bedrock of aviation safety: the pilot in command.
Convenience at what price?
Steve Jobs once famously said the people “who are crazy enough to think that they can change the world, are the ones who do.” Innovators are impatient with caution, and the retail drone industry is driven by innovators. However, when successful innovation entails risk to human health and welfare, caution is not only warranted, it is essential.
As we move toward the seeming inevitability of retail drone delivery, safety must continue to be of paramount concern.
The issuance of Part 107 illustrates that the FAA remains committed to this principle, but the economics of retail drone operations will likely test that commitment going forward.
William H. Walsh is co-chair of the aviation industry practice group for the law firm Cozen O’Connor in Seattle and a former U.S. Air Force officer. He can be reached at firstname.lastname@example.org and 206-224-1296.