On July 1, 2020 the FAA released its V1.0 vision for an urban air mobility (UAM) airspace access concept [link]. Essentially, the FAA is proposing to have a series of “aerodromes” connected via “corridors” in which UAM aircraft can operate largely separately from the rest of the national airspace (NAS). Within those corridors, UAM aircraft would be in contact with a Provider of Services for UAM (PSU) which would subsequently be part of a network of PSUs, facilitating the data channels necessary for deconfliction and communication with and between UAM aircraft. Also within those corridors, operations would be governed by Community Based Rules (CBRs), which the FAA intends to approve but not to write. The whole system is intended to accommodate a progression through “crawl”, “walk”, and “run” phases in which the industry is increasing both the frequency (or “tempo”) of operations and the extent of autonomy that is present on the aircraft.

Schematic of aerodromes and corridors from the FAA UAM CONOPs v1.0

While I can appreciate some of the challenges that this CONOPs is attempting to address, I can’t help but think that it’s heading down a path that will be detrimental to both the UAM industry and the communities it intends to serve. While I am very much involved in many of the industry activities that will be necessary to make this whole thing work (GAMA’s EVTOL Committee, ASTM F44 on General Aviation, and CAMI, to name a few) this isn’t intended to be the official position of any of those groups or any of the companies I work with. I also want to, right up front, thank the FAA for being willing to lean into this space in any way and very much see them as a collaborator in making UAM happen; the published document is a first version and as such it’s a great place to start the conversation.

On the fundamental assumptions:

First, in the background information, it is implied that UAM is a solution to traffic congestion and that it will “provide an alternative to intermodal transportation”. I don’t believe either of these to be true. History has shown that increasing capacity does not in fact reduce traffic congestion, it simply creates more demand. As such, I do not believe that UAM is a solution to congestion: rather, it is a tool by which the third dimension can be added to a community’s transportation landscape, and to do so well will require that it be integrated with other modes of transportation, including mass transit. The language in the CONOPs is concerning: developing a system in isolation from and without the ability to seamlessly connect to, existing transportation options will hinder the utility of UAM.
Another fundamental assumption of the document is that all UAM aircraft will have a human pilot in command (PIC) on board each aircraft. While the debate about whether this is technologically necessary or indeed safer than a “direct to autonomy” route is out of scope of my purposes here, it is presented as if this is the course that “industry manufacturers and operators” prefer. This highlights a chicken-and-egg problem in the industry: the industry prefers this in large part because the FAA has not presented an efficient path by which anything else can be done. While it may well be the best course of action (it may also not be), it should be presented simply as what the FAA intends to support, not presumed to be the FAA being responsive to industry.
One piece that seems to be missing is vehicle-to-vehicle (V2V) communication. This is something that has been in discussion in industry for some time and could enable faster and higher bandwidth communication between aircraft operating in proximity to each other, whether those aircraft have a human pilot on board or not. With simple datacomm protocols in place, four-dimensional (4D) trajectories can be shared, deconfliction and sequencing can be automated, and emergency landing locations reserved in real time. One of the requirements for aircraft to access a corridor (which it’s clear there will be) can be appropriate equippage for V2V digital data communication. More on this in the discussion of CBRs below.

On the corridors themselves:

The proposal bears a striking (and acknowledged) resemblance to the existing helicopter routes concept, with the key difference being that inside the corridors (and despite a stated desire to the contrary), it appears that the aircraft are cut off from other NAS activities and ATC services.
The geographic determination of the aerodrome locations and the corridors themselves is a question that must have a high level of input from local authorities and the communities in which the aircraft will be operated. While it is clearly stated in the document, and I support the idea, that the FAA has preemption over all airspace, the document is silent on the role that local land use zoning and other tools might play in ensuring that appropriate effort be made to achieve the maximum balance of economic and community benefit, utility, adverse impact mitigation, transit integration, and social equity for these corridors. There needs to be a clear process for local authorities to be involved in a meaningful way in the FAA’s approval of corridors.
Once the designated UAM corridors are defined, it wasn’t immediately clear to me how the FAA intends to protect them from non-cooperative traffic. As the corridors can cut through any type of airspace, it is possible that a wide variety of other traffic could be operating in close proximity to the boundaries of the corridors. The document mentions crossing traffic, but doesn’t expand on how that is going to be handled. For the full potential of the system as described to be realized this topic will need to be fleshed out and precautions put in place. Simply marking the corridors on sectional charts or treating them like TFRs (or similar) does not, in my opinion, provide the level of airspace protection that is needed. Without strong airspace protection of the corridors, it’s not clear that their isolation from the mainstream ATC services isn’t simply a liability.
I will say that I don’t actually have a problem with the idea of corridors, particularly in the early stages of the UAM industry, though I think a critical eye should be kept open to the possibility that it may become obsolete or unnecessary as the UAM industry and aviation in general evolves. It is even possible in the future that the operational practices and CBRs that are developed and demonstrated within the corridors could eventually become more widely applied (but this is out of scope for now).

On the Community Based Rules (CBRs):

As mentioned above, the concept of distributed deconfliction and prioritization between aircraft within a corridor is a powerful one that appears to be missing from the document. The industry has been having conversations for some time about how this could be accomplished, what the “rules of the road” could be, and how they might be distilled down to the most essential data and algorithms necessary to ensure safe and efficient flight in congested airspace.
Distributed self-deconfliction and sequencing allows aircraft to avoid each other, to self-sequence into landing sites, and ensures that each aircraft always has a dynamic and unique reservation on an alternate emergency landing site should an off-nominal situation arise. V2V digital datacomm need not be large nor power intensive. Bandwidth requirements are trivial for safety critical messages and could potentially fit into the 5.9GHz spectrum already allocated to DSRC for vehicle safety by the FCC. Most importantly, the system would be inherently more scalable than what is proposed. The largest challenge would simply be agreeing on communications protocols for messaging and ad hoc networking -- which could be done either in RTCA or ASTM. Most importantly, it could reduce the criticality of the PSU construct.
All that being said, CBRs that are based on distributed self-deconfliction can increase safety and lay the groundwork for operations that do not require a PIC on board the aircraft. CBRs that are not written with this in mind will not set the industry up for the same level of long term success.
In my opinion, the best process by which to arrive at a set of CBRs is to leverage the efforts of GAMA’s UAM-focused EPIC committees, which include EVTOL, DataComm, and Infrastructure, along with the ASTM F44 Committee on General Aviation Aircraft, which already maintains a relevant library of means of compliance standards for aircraft certification and has the processes and regulator participation in place to facilitate rapid and robust standards development. As I’ll discuss next, including clear requirements for PSUs, if they are ultimately deemed necessary, should be part of this process as well. Drafting these CBRs should be a priority for industry.

The role of the PSU:

I can appreciate that there may be a hesitation on the part of the FAA to sign up for managing an increase in airspace usage on the level of what is expected from later-stage UAM operations. Particularly in a resource-constrained environment, the idea of privatization may seem like the only viable option. Privatization is exactly what the PSU system is, however, it is one that carries real risks to the ideals of not only safety but equitable access to the airspace.
From a safety standpoint, adding entities into the communication chain is simply another way of adding potential failure points. Each aircraft could, theoretically, be in contact with a unique PSU. Each of those PSUs then has to be in contact with each other on the “PSU Network”. FAA ATC for the broader NAS is then connected to the whole thing as well. This is going in the opposite direction of increasing system robustness through distributed deconfliction. Each of those communications hand-offs is an opportunity for data corruption or lost link. If this is being done on the assumption that the data sharing necessary to do distributed deconfliction is not available today, then that assumption needs to be reevaluated with industry. Another strike against the safety of the PSU model is the feasibility of ensuring that each PSU is actually following the applicable requirements and operating in a manner that upholds the expected level of rigor in their operations. The FAA may well be trading an ATC responsibility for an auditing and oversight responsibility, potentially spending just as many resources on enforcement and what amounts to babysitting as they would spend on overseeing a largely self-sufficient distributed system within the corridors, all at the potential expense of safety and, as we’ll look at next, equity.
The PSU construct presents a risk to equitable access to the NAS. By giving PSUs the ability to provide demand capacity balancing (DCB), or basically the ability to to tell an aircraft that they can’t fly in the corridor, we’re creating a system that could be eerily similar to a lack of net neutrality. In other words, those that pay more can fly more. This is in stark contrast to our current airspace model, where landing fees may be charged by airports (or vertiports) but the air itself is a public resource open equally to all.
There is also an interesting overlap between the PSU construct and the desire from some municipalities to have real-time data for transportation usage and the ability to do real-time throttling of the transportation system’s capacity. While this may seem like a good thing, and it is desirable to have communities be empowered to impose reasonable restrictions on UAM operations, for example, based on time of day and the audible impact of the aircraft, it is important that a dynamic ability to limit access to the corridors doesn’t undermine the reliability of UAM as a daily mode of transportation. There will likely be situations in which dynamic throttling of corridor operations makes sense - for instance in the case of an emergency like a structure or wildlands fire - if there is too much variability UAM may gain a reputation for being unreliable. The guidelines for when and how corridor access may be limited must be clearly stated and consistently implemented across PSUs and jurisdictions.
A note on the PSU-USS comparison: many of these same concerns can also be expressed for the UAS Service Supplier (USS), and I am not a huge proponent of that either, however the associated risks are magnified with the PSU concept. Among other things contributing to making the PSU concept even more concerning than the USS concept: the aircraft are larger, operating in closer proximity to other traffic and other parts of the NAS, and of course, they are carrying people.
Crafting good standards to which the PSUs will be held can mitigate some of these concerns but does not eliminate them. Fundamentally, splitting authority from responsibility, as is being proposed here, is something that has always been a mistake in my personal experience. It creates inefficiencies, failure modes, and frustrations that are avoided by maintaining that integrity. I would strongly suggest reevaluating the cost-benefit analysis that led to the desire to create the PSU construct. By leveraging V2V digital datacomm along with the protocols and “rules of the road” to be defined in CBRs, deconfliction and self-sequencing can be done within these corridors with minimal direct FAA involvement.

To sum up, I am heartened to see the FAA devoting significant thought and resources to figuring out a path to initial UAM operations that goes beyond simply trying to squeeze them into the existing ATC construct. I agree with the assertion of the FAA’s preeminence in its authority over airspace but think that there needs to be more thought on how this interacts with local jurisdiction over land use and community integration considerations. I appreciate the awareness that this is an evolving industry and that the extent of autonomy that is in use in these aircraft, as well as the frequency of their operations, is going to be higher than in conventional GA and will continue to increase with time.

I am concerned that the corridors may not be appropriately defined or adequately protected from non-cooperative traffic. I am concerned that the creation of multiple private entities all involved in the communications and data network necessary to safely and effectively operate these aircraft will decrease safety, increase unanticipated costs for the FAA, and decrease equity of access for the NAS. It also potentially undermines equitable airspace access. I would much prefer an investment in V2V data protocols and distributed deconfliction and sequencing that can be done directly between the aircraft in the corridors with a procedure to interface directly with the broader ATC system when forced to diverge from the corridor. The existing industry efforts in GAMA and ASTM F44 (with collaboration with F38 on Unmanned Aircraft Systems) are well-suited to support the development of the necessary standards.

I am looking forward to continuing the conversation between the FAA and industry, wearing a variety of different hats, as we all drive to a constructive, safe, durable, and effective solution that meets the needs of all of the stakeholders within existing constraints.