‘On My Tombstone, It Should Say Connected the World, Not Created Orbital Debris’

by
Greg Wyler’s audacious vision for connecting the world via a 720-satellite constellation, rooted in part by time spent in Africa hooking up remote villages with terrestrial links, has no successful precedent.

But $500 million of initial investment capital and a list of ground-floor partners that includes heavyweights Qualcomm, Coca-Cola, Totalplay Communications and Bharti Enterprises go a long way toward building credibility. So much so that space hardware and service providers have been actively gearing up for the day that OneWeb comes knocking.

For some, notably satellite manufacturer Airbus Defence and Space and launch service providers Arianespace and Virgin Galactic, that day has already come.

But OneWeb, which hopes to begin launching satellites in 2017, still has a long way to go and much to prove. The company has yet to finalize U.S.-based satellite manufacturing arrangements, there are nagging questions about radio frequency interference and $500 million isn’t nearly enough to get the constellation system built and deployed in the first place.

Wyler, who built his satellite industry chops by founding the O3b Communications venture and getting it off the ground, addressed these and other matters in a recent conversation with SpaceNews correspondent Debra Werner.


Are the OneWeb satellites under construction?

The subcomponents, we call them the long pole in the tent, are either in development or getting ready to be manufactured. It is all based on a schedule. It is a pretty in-depth and involved schedule of hundreds and hundreds and hundreds of components that are coming together; some obviously will be shipped to the factory where the satellites are going to be manufactured. But most of the long-pole tent components are already under contract or underway.

What is the likelihood that the initial OneWeb launches will slip 2018?

There’s always a chance. This is space. We expect Arianespace to be able to deliver on time. The question is will we be ready? We hope to be ready. My experience with Arianespace in the past has been that they launch when they say they are going to launch. I’ve had very good experience with them in the past on the timing.

In June you announced a lineup of contractors plus a couple of service partners in India and Mexico and $500 million in equity financing. Where are you now in terms of financing and in terms of creating a joint venture with Airbus on satellite production?

We announced our fundraising first round back in June and I’m really excited and happy about the investors that gathered around the table to support the mission and support the vision. And the investors were across the whole gamut, some that were participating on the manufacturing side, some that were participating just as financiers, some that were participating as customers — they will be resellers in their countries. It’s really great to see the whole ecosystem come together. Right now for us, we’ve raised that money. We are not raising money right now. We are finishing off some other financing aspects of the system, locking down a whole bunch of things. Probably in maybe 18 months, we would be going to get the next round. It will be one more round.

Will it be a round of similar size?

About the same size.

Will certain financing milestones need to be reached before you start building satellites?

No, there is no milestone financing. Of course, when we do export credit, the different debt structures may have some milestones, but we don’t have any milestones right now.

Speaking of export credit, are you counting on support from the relevant agencies in France, Canada and the United States?

We would like to have everybody, but we’ll see how that works out. So far we found that we have a tremendous amount of support. All lights are green. Everything is going well. We are on schedule.

In the United States it looks like the Export-Import Bank is going to be back in service after a hiatus, but likely will have some backlog to work off. Will that slow you down in any way?

I don’t know yet. It is just too early to tell.

When can we expect to see the opening of the OneWeb satellite production facility in the United States?

We look forward to announcing something.

Word has it that the plant will be in Florida. Can you confirm that?

We look forward to announcing something.

There is a lot of concern that the OneWeb constellation will cause unacceptable Ku-band interference to other satellites in the equatorial regions. What assurances can you offer that this will not happen?

That is a great question and fortunately the EPFD [equivalent power flux-density] system, the EPFD regulations, have been well formulated. There is software that will be coming out from the International Telecommunication Union and that software will give us a thumbs up/thumbs down on the Ku-band EPFD interference — whether you meet that EPFD requirement. And it is with mathematical certainty, or you don’t meet it. Once the ITU releases that software, then everybody will have to release the EPFD masks to run through their software. That will give them all the assurances that are required.

Space hardware and services provider MDA of Canada recently said it is part of the OneWeb contracting team. Could you elaborate on MDA’s role?

Sure. MDA is a great team and they’ve been supporting us in quite a number of aspects of the payload. Long term, exactly what that role will morph into, we’ll see.

Is your entire manufacturing team in place?

No, because it’s too early to have the whole manufacturing team, but certainly the core of it is put together and there are a lot of people in Toulouse, France, right now working on stuff. The unique part, the way we approach a system of this size, is we have a lot of partners who each have tremendous capabilities, like Qualcomm or Hughes. There are teams of people that already exist in these very large companies that get assembled or joined together to solve our problems as part of the whole. We can put together fairly large teams very quickly without having to hire or go through that whole process and move people and the like. So we’ve actually been able to spool up certain key technologies quite fast.

Was there a specific moment when you adopted this vision of Internet access for everyone?

No, there was not a moment. It was a thing that happened over time, after seeing the impact of bringing Internet to individuals and villages and sort of coming up through the ranks. I got a chance to crawl up the side of a hut in Africa and put up a wireless access point, lots of them, and actually run the wires down the sides of the houses and bring connectivity to communities. I saw the impact it had on individuals. I hired people and worked with them, people who had not really had access to the world’s information. You could see that they changed, for the better. A number of them, maybe 10 percent, became so enthralled with it that it became a part of their life.

When did satellites enter into the equation?

I did not look at the problem from the perspective of putting up satellites. In fact, O3b originally was about trying to run fiber and it just didn’t work. It was just too hard. I couldn’t get through all the regulatory aspects. I didn’t see a good path to building something there. It could be done and it is being done but it is going to take a long time to run fiber across some very remote areas and then maintain that fiber.

You’ve said that space is not the most challenging part of connecting everyone. What is?

For connecting everyone, it is having a technology that makes economic sense and having a business model that makes economic sense. For the project, the challenging parts of course are execution on a timeline. We’ve got a lot of pieces coming together and we have to just methodically work our way through them first. Probably the individual hardest part are the chips in the user terminals, the chips that actually manage all the handovers from beam to beam and the rates and speeds and the amount of megahertz that we’re dealing with. This is some very, very sophisticated technology. Fortunately, we have great partners. Qualcomm understands that problem really, really well since their technology is in 95 percent or so of cellphones. They build hundreds of millions of chips a quarter that are devoted to figuring out how to do handovers in just really wacky cases.

Your initial constellation will be launched on Europeanized versions of Russia’s Soyuz rocket. How many will be launched at a time?

The ultimate answer is that is not yet fully determined because we may actually vary it for a whole bunch of reasons. But you can put about mid-30s, 32 or so, in a Soyuz. There are a whole slew of reasons why we may put more or less.

Such as?

We may want to do some larger satellites for certain experimental purposes, in which case the mass will go up and the number of satellites that go into the rockets may be less. We may need to fill only a certain portion of the system and it doesn’t make sense to launch more satellites into the orbit that we are going to be launching in. There are a variety of things that would come up that will allow us to put more or less onto the Soyuz. But it will probably nominally average out to about 32.

How does Virgin Galactic’s LauncherOne vehicle fit into the overall scheme?

Virgin’s capability is speed to space. We will use it to refresh and refill the constellation at points of failure. So if failures happen, which they will, we will be able to throw satellites, launch them directly into the plane very quickly. We have room for failure. We expect certain levels of failure. It’s just part of the business.

What are you doing to mitigate the potential of creating orbital debris?

It’s actually something that our team takes very seriously. The question is, how do you design a constellation that has appropriate miss distances? How close do your satellites come together? And how close do they come together in all sorts of failure modes and all sorts of natural conditions, solar winds and the like, that may perturb the orbits? You can’t have orbits that if the inclination is perturbed by half a degree or one-quarter of a degree you’ll have a collision. So you really have to design these things very, very carefully. So we have designed our system with the failure modes necessary to reach a level of safety from an orbital debris standpoint. We’ve designed a system with the orbits such that they have enough miss distance. That’s super important. On my tombstone, it should say “Connected the world,” not “Created orbital debris.”

And of course you plan to deorbit your satellites within five years of their end of service.

Absolutely. You have varying degrees of failure modes for the different systems within the satellite. For instance, if a transistor of some other component fails, we bring it down. Whatever. We’ll put another one up. But you can’t have the failure mode aspects if the deorbit systems fail. So you want to have a lot of reliability in that deorbit system.