Amit Ganjoo, founder and CEO of ANRA Technologies, discussed unmanned traffic management with Avionics.
TYSON’S CORNER, Va. — As regulators, air navigation service providers and private companies continue work toward integrating unmanned traffic into the global airspace, Avionics International visited the offices of ANRA Technologies, a provider of airspace management services and workflow orchestration, to discuss unmanned traffic management with founder and CEO Amit Ganjoo.
Ganjoo has been involved in UTM development since 2015, and his company has participated in every NASA and FAA discussion since then. ANRA was recently awarded a contract to assist the nation of India with UTM integration.
How did you get involved in the UTM space and come to found ANRA Technologies?
My personal background is in communication, aviation and robotics. I used to be the principal architect for a company called Ericsson, and aviation has been my passion since I was young. I'm a pilot as well. After that, I built my own experimental aircraft, and based on my background I was recruited to work for the U.S. Department of Defense on communications and drone programs. I won’t get into details there, but drones have been in the defense space for a long time, so I saw the transition to commercial use happening and realized that the set of challenges that we experienced in defense would later be faced in crossing over to the commercial side.
In early 2015, we found some patterns, built some quick and dirty prototypes and took it to an event in D.C., where we got the “coolest technology award.” That made me think, hey, there’s something here, and that’s when I left DOD to start something in the commercial space. A few weeks later, someone at MITRE connected me with the principal scientist at NASA working on UTM, Parimal Kopardekar, who was looking to research what we were already working on. So we were one of the first companies to enter into a collaborate research agreement under the Space Act in 2015, and we are probably the only company that has been at all the NASA national campaigns, from TCL-1, 2, 3, to 4, and that was our entry into the UTM space.
I know you’re very active in industry standards-making bodies related to UTM; can you tell me more about your activity there?
I was involved in standards when I was on the communications side, too; I used to be on the FCC’s technical advisory council for 5G and IoT, which included drones. I’ve worked with ASTM on standards for UTM and remote identification for drones, and I’m on the board for the Global UTM Association, GUTMA, as well.
Tell me more about GUTMA. How did it come about, what is its mission?
You have to understand, UTM started as a concept primarily in the U.S., as a NASA research project, and then it took everyone by storm. So Europe and the rest of the world was lagging behind on this, and the restriction with the NASA project is only U.S. companies can collaborate. So a lot of foreign companies and countries decided to start their own UTM projects — parallel efforts, such as the U-space project in Europe.
So, there were discussions about the need to harmonize UTM at a global level, and since NASA cannot do it, an industry association should do it. We formed GUTMA about four and a half years ago, and initially many of the member organizations were from the U.S., but now we have expanded to 68 members ranging from ANSPs, regulators to industry participants like us.
How would you characterize the current status of global UTM efforts; has there been a lot of divergence in how different countries are approaching unmanned traffic integration?
That’s what has happened right now, because initially there was this paradigm of, “it wasn’t invented here, so it’s not good,” with many countries trying to go through the same iterations that NASA started in 2015. Many of them are trying to do their own one-off thing and it’s still an uphill battle, but we are educating folks that, hey, it’s okay to try new things which are specific, but let’s build on top of lessons learned. So that’s where the harmonization comes in. And we have gotten a lot of them to change their minds and come to an agreement on a base level of standards, and then you may have some variants or add-on optional things that are country or regional-specific.
Let’s talk about infrastructure. What is required to enable UTM in cities, suburbs and various areas — and how does that compare to what currently exists?
That is a tricky question when it comes to urban air mobility, but not as tricky for UTM. The original concept for UTM was low-altitude airspace, below 400 feet, using tracking and strategic and tactical de-confliction, and if there are off-nominals or non-conforming aircraft the system will notify everyone else who is impacted so they can take preventive action.
To accomplish that, from an infrastructure point of view, I need a link back from either the GCS [ground control station] or the UAS, sending me the aircraft’s telemetrics and GPS so I can track you. It’s relatively straightforward when you do this for visual line of sight flights, but when you go to BVLOS — that’s where the detect-and-avoid, surveillance data and all those things come into play.
Right now, the density isn’t there, so I can pretty much put down surveillance infrastructure and combine that with onboard capabilities. A lot of pilots like BNSF [Railway] are putting down their own small radars, pull in that data, and maybe put onboard radars and detect-and-avoid capabilities to build the safety case for the FAA.
Thinking ahead to, let’s say, widespread drone delivery across the suburbs — that will have to be BVLOS or it won’t be scalable. Some companies say the best way to achieve this is installing small radars on traffic lights and telephone poles across a county. Is that the solution?
It will be very use-case specific, and it’s about the safety threshold and the safety case. If I’m operating in the middle of Nebraska or Iowa, I don’t need any ground infrastructure as long as I have decent onboard capabilities. Now, if I am flying in downtown D.C. or downtown Manhattan, in addition to two communications links in case of failure and onboard sense-and-avoid in case I fly into an urban canyon and my GPS goes haywire, it will be helpful to have surveillance capability, and that’s where ground-based infrastructure comes in handy, so that I can have localized data keeping me informed. But it’s not ubiquitous — I don’t need it in all cases.
If you look in realistic terms, in order for UTM to operate, all I need to get from the drone is the telemetrics. That is in kilobits; I can do that on a 3G, 4G network; I don’t need high bandwidth. That would be useful for other drone applications — live streaming, pulling in tons of LIDAR or other data — but not for traffic management. It’s not a bandwidth issue.
So it’s not a bandwidth issue, it’s a reliability issue?
If your paradigm is that you’re always going to be connected and that’s how your system is going to track aircraft and notify everyone else, then reliability becomes a key metric. But if I’m flying a BVLOS operation and there’s a dead zone that reaches a few hundred meters or maybe a half mile, that may be okay. If I’m flying a waypoint-based mission and I lose my link for ten, fifteen seconds, but I know I’ll have coverage at the next waypoint and get the link back, as long as you have enough confidence in your aircraft to maintain its trajectory, then the system might be okay. It’s very specific to your safety case, your equipage and your aircraft.
So in a city with dead zones, you can do drone operations at a small scale, but for large scale — or especially for urban air mobility — we will need infrastructure upgrades. That will require weather sensors deployed for data on urban canyons, and surveillance feeds coming in to ensure the separation of aircraft, especially when the density increases.
How many of these sensors will we need; how densely-packed?
It depends what resolution you want. Typically, if you look at current weather, it’s based on ground stations and their resolution is a few miles. But when you’re flying a small UAS, you need it in feet, not miles. So you might need a weather sensor every 100 or 200 feet; maybe every other building, depending on how far apart they are. But just like with cell tower planning, you can divide the area into a grid and place them based on what resolution you want to achieve.
What’s the current status of discussions around how to connect UTMs with air traffic control systems, and how do you see that playing out?
So right now, we have the Low-Altitude Authorization and Notification Capability, or LAANC, around airports, but that system doesn’t actually track anything. It just shows air traffic controllers flight plans, or intent to fly.
In true UTM sense, air traffic controllers need to be able to track as well, to see that traffic if it’s going to impact anything. That’s where different countries around the world are at different stages and are taking different approaches. The UK, for example, has decided its airspace is so dense and such a small area that they have to approach ATC and UTM together or it’s going to be a nightmare. Other countries have a lot of land mass and areas where the air traffic control system doesn’t care if someone is flying a drone unless it’s near an airport.
But true UTM integration would mean more than just adding a data feed to ATC displays; it means controllers can make decisions and filter based on what traffic they want to see, and there’s two-way communication.
For example, in some of the project we are working on, including India, they want all the ATC feed to come into the UTM platform and the UTM feed to be available there, so it’s full two-way visibility, combined with filtering capabilities to view flights relevant to a specific area. Everyone has a different safety threshold and idea of what they want to see.
Speaking of two-way communication, Jay Merkle, head of the FAA’s UAS integration office, recently mentioned the agency will put out an RFI looking for industry solutions on how manned aviation can take advantage of information on where drones are. How do you think this should be implemented?
If you look at UTM in terms of a network solution, a UAS service supplier will have all the data. We have done remote ID demonstrations in California, New York and in Switzerland — with Airmap and Swiss FOCA — where we could see their traffic, they could see my traffic and our clients could pull it up. So that data is available and we have that capability.
So the simplest, low-cost solution is to just to pull the data from the UAS service provider?
I would say relatively low cost since for the UAS Service Supplier to have the data, it is expected that the operator, will have network connectivity which in turn means that the UAS operator will incur some data service costs. In some cases, if the drone already supports a wifi hardware module like many current drones on the market, it might just be a firmware update. So again, a low cost option. As you can tell, there are pros and cons with each of the options.
What else can you share about ANRA’s contract with the government of India and that nation’s approach to UTM?
The India project was kicked off around August last year, give or take, and this phase is a two-year project, though it may be prolonged a bit. They’re still figuring things out from a regulatory point of view, but there have different zones — zones where you cannot fly at all, zones where you can fly, and zones where you need waivers. Initially, there’s a huge part of the country that’s not going to be available for drone operations at all, but over time those zones will change.
We are helping to educate the government, because they started doing this in a vacuum on their own; remember what I said about UTM fragmentation around the world. So they put some things in their regulations which are not necessarily ideal from a future-proofing and scalability point of view. We’re working with them hand-in-hand to guide them and hopefully impact future regulations.
This interview has been edited and condensed for clarity.