Roundtable: The Rapid Evolution of In-flight Connectivity

In-flight connectivity is a rapidly evolving field where connectivity is a must-have, not a nice-to-have. Once, passengers were hoping for the ability to text, and now they want offices in the sky. Reliability and speed have become commonplace, rather than a target. AIN brought together experts to discuss how the industry is meeting the demands of these changes and what to expect in the future. Viasat sponsored this roundtable.

The Participants

Claudio D’Amico — vice president of strategic market engagement, Business Aviation for Viasat

Claudio D’Amico has served with Viasat since January 2014, initially as program director. He began his career in 2001 with Embraer as a market analyst and also held management roles with Goodrich and later UTC Aerospace Systems MRO Brasil. A global communications provider, Viasat has served the business aviation sector for more than 30 years and has some 5,000 jets in its network. Viasat, which in 2023 acquired satellite specialist Inmarsat, offers an array of services from L-band and SwiftBroadband to Ku-band and Ka-band.

Jay Heublein — president of Flexjet Technical Services 

Jay Heublein has served with Flexjet and affiliated and predecessor companies for more than two decades, also holding leadership roles with Flight Options and Nextant Aerospace. He has further held a senior sales position at CitationShares. Fractional ownership provider Flexjet is one of the largest business aviation fleet operators globally. The Cleveland-based company is also an integrator of Starlink and has been converting its 400 aircraft to the service. Involved in in-flight connectivity for decades, Flexjet first introduced air-to-ground [ATG] technology in the 1990s and has worked with an array of connectivity technologies.

Jason Wissink—president of services and connectivity for Honeywell

Jason Wissink has served with Honeywell for more than 17 years, beginning as a senior program analyst in 2007. Also a former software engineer with Intercomp, Wissink has been involved with the connectivity side since Honeywell introduced its JetWave connectivity solution more than a decade ago, working with Inmarsat. The aerospace giant now has more than 1,000 installations and is working on its next-generation product.

Michael Schrage—research fellow, MIT Sloan School of Management

Michael Schrage, a research fellow with the Sloan School’s Center for Digital Business, is also a visiting fellow at Imperial College’s [London] Innovation and Entrepreneurship program and has authored several books on the role of collaborative tools and technologies in enabling innovation. He has further served in consulting and advisory capacities with Viasat, Microsoft, Procter & Gamble, British Telecom, BP, Siemens, Embraer, Google, iRise, and the Office of Net Assessment, among others. Schrage has delved into the shifting key performance indicators (KPIs) of network design and connectivity when measured with user experience in commercial and business aviation.

The Discussion

On the Business Aviation In-flight Connectivity Landscape

Claudio D’Amico

There’s definitely a change in our industry. We went from being able to send a text on an airplane, and that was good enough 20 years ago, to using that aircraft as an extension of our office or living room. That’s the expectation when you’re flying on these business jets, but at the same time, the technology of the aircraft has improved significantly. We have airplanes flying longer distances, and I think that also has a big impact on connectivity for our industry. [New] jets are equipped with connectivity as a must-have solution. It went from a nice-to-have to a must-have.

[There is a] network of providers. Over the years, [they] saw this market as an attractive market, some successfully, some unsuccessfully. There’s consolidation happening. It happens in every industry that is maturing. You have innovators, then some consolidation, and we’re seeing some of that.

The demographics of our passengers are changing. At one point, speeds were very important. We participated in that process. But now, with this maturing in the industry, we’re seeing experiences being more important. And how do you characterize that? That’s how the whole customer base is evolving.

Obviously, we’re seeing LEO [low-earth-orbit] entrants. From Viasat’s perspective, competition is good. It forces us to continue to innovate and bring better solutions to our customer base.

We think that a multi-orbit solution backed by high-capacity and flexible satellites is where you can continue to innovate and drive improvements in terms of experience. The folks who focus on customer problems and at the same time continue to evolve and deliver capacity in a very efficient way [will prosper].

We are continuing to expand our network. We have some very high-capacity satellites being launched and integrated into our network. Viasat entered service with ViaSat-3, a high-capacity satellite, in the summer. We have more of those satellites coming. Those satellites bring an immense amount of capacity to keep us ahead of the game.

We’re repurposing some of our assets and bringing five times the capacity over North America for our GX customers. We’re not just focused on a GEO solution or geostationary orbit for our satellites. We have our GX-10 satellites coming online next year, and those are high-elliptical-orbit satellites that provide coverage over the North Pole. It’s the first step in terms of multi-orbit for us. The backbone for us is being able to have these high-capacity GEO satellites and focus that bandwidth over areas with a lot of demand.

Jason Wissink

We’re seeing a combination of multiple things that are expanding the choices that operators have. The established providers like Viasat are adding significant amounts of capacity. Our goal is to make sure that the available equipment—not just available on the shelf but certified on the right airframes—has access to the latest and greatest capacity. We have to stay very close with companies like Viasat to understand those roadmaps, timing, modems, waveforms, all that stuff, so that when the capacity becomes available, people can get access to it right away.

On top of that, there are a lot of new entrants and disruptors. Starlink is one of them, and they’re driving a change into the market, in terms of what experience should be expected. As a manufacturer, we have a combination of people that we’re partnered with, and then some folks are more vertically integrated that we still need to find a way to work with, because you end up with aircraft that have multiple systems installed. What passengers want is a system that works great and provides a great experience.

We’re just trying to make sure we’re doing everything we can to have equipment available that is certified, available, and can access the latest capacity, and also is set up to be integrated with other systems.

We’re seeing more and more [installation of systems that connect to different networks]. It’s already happening in the airlines. Multi-orbit and business aviation is a bit more of a challenge from an equipment perspective, because you’re limited in terms of the size of the fuselage. You can’t put massive antennas [on them]. But we do have access to the tail. You can put smaller ESAs [electronically steered antennas] on the fuselage. If you find a way to integrate those types of systems well and provide a very good experience, you can get beyond just redundancy. Redundancy is great, but these systems are pretty reliable now, and you can get to a point where you’re running them both at the same time and providing an even better experience than just one or the other.

Michael Schrage

Let’s not confuse greater capacity with capability and how that capability is experienced. This is where the Honeywells of the world are going to play an enormous role. Jason talked about systems integration and the user—whether the user is a pilot or in the back of the plane, the user doesn’t want to deal with that kind of complexity. They just want the thing to work. What I think is going to become very important in the business space and in the commercial space—what I’m about to say is classic UX [user experience] design—but what are the personae of the users, and what are the edge cases that will test the capability of the system?

One of the real issues for Starlink, for Viasat, for Honeywell, for Collins, is if you’ve got the children of [the owners] on a jet, and they are influencers on Twitch, how well can they do stuff on Twitch as opposed to just stream a Netflix movie? What are the edge cases? What kind of latencies are acceptable?

These are the big issues, and that raises one last issue. I think metadata matters. I think the intent of the user matters. And I think the Honeywells and the Viasats are going to be doing packets that are metadata-labeled as “this is a game, this is a Zoom.” This is a high-priority issue.

I think there’s going to be more intelligence in the system, and capacity is not just going to be about speed or greater speeds; it’s going to be about facilitating intelligent continuity for the user.

Jay Heublein

We all run our businesses using [key performance indicator] KPI data. [Last] Sunday, we had 380 live legs. The only negative KPIs that I saw, somebody had bad catering on one flight, and on three flights that had legacy connectivity equipment, they complained that the connectivity didn’t work. Think about how complicated it is to run [those] flights, all the logistics, all the operations that go into that. We can run a perfect flight. The airplane could be perfect. The flight can be on time. And now we’re down to people are focused on whether or not their in-flight connectivity experience mirrored what they expect in their office or in their homes. The expectation is that anything I can do in my home, in my office, I should be able to do in my aircraft.

Four to five years ago, from our perspective, we were struggling. It was interesting because the early generation stuff was great. We were all happy when we could do basic text and emails back in the ’90s, and then, almost immediately, “I can’t stream. I can’t host video conferencing.”

It’s amazing how fast the expectations have evolved. Five years ago, we were challenged with the lack of consistency in in-flight connectivity. There’s been great progress in technology. And Ku was a big step up from ATG, and then Ka got even better. But if you have a less-than-perfect connectivity experience, our customers think of the flight as a failure.

We thought about getting into the connectivity space. We put a fund together. We were thinking about buying a connectivity solution. We looked at all the different options. And during that diligence, we came across Starlink.

We started looking at it, and we focused on the things that create challenges for us. When you’re flying 400 aircraft, Ka and Ku were a huge step forward. It’s very complicated equipment, though. The thing that we liked about Starlink was this idea that if we get away from moving [antenna] parts, anybody who manages airplanes would tell you there’s an advantage in that.

We realized two to three years ago that they were going to be able to build a phased-array antenna [or ESA]. We thought that was a big step forward, because we got away from moving parts. Even though it had gotten dramatically better over the decades that we’ve been involved with connectivity, there were still high failure, high-priced equipment that we had to deal with. If you don’t have a perfect in-flight connectivity experience, people are upset. So we decided to go all-in on Starlink.

Then, just like we do with a lot of our businesses, when we try and solve for something with Flexjet, we look at monetizing it in different solutions. So, we’re an integrator for Starlink. I think [we’ve developed] about 20 STCs [supplemental type certificates], and in a year, we delivered almost 800 kits.

It’s going to take two years to integrate Starlink on all 400 [aircraft]. We’re about a year into it. Now the complaint I get is, “I’m still flying in an ATG-equipped airplane,” and the experience isn’t the same. That’s our challenge. In the last couple of years, we’ve crossed the reliability threshold. In the 110, 120 airplanes that I have flying with Starlink, I’m aware of two mechanical issues, which, relatively speaking, is incredibly low compared to what we’ve dealt with 10, 15 years ago with in-flight connectivity.

On Bandwidth

Jay Heublein

When we talk about KPIs, we woke up one day and the average number of devices that are connected went from maybe six to eight to somewhere north of 38 to 40 on an average business jet flight, even though there are only three passengers in the back of the airplane. It’s gotten incredibly more challenging to meet expectations, and that’s what the new equipment is doing, and that’s what we’re excited about.

Michael Schrage

What you’re describing is literally an order of magnitude more devices. How are you using metadata? Do you have to become a systems integrator in the air to manage these sorts of things? And do you have to create a new level or layer of the stack to coordinate and orchestrate what’s going on? Because you and I both know that a certain portion of your flyers are having ChatGPT and Claude [AI chat] conversations, and at the same time, their kids are streaming things. How are you addressing the orchestration, coordination, and intelligence challenge of that? Because it’s 10X more now than it was five years ago.

Jay Heublein

Think about 14 to 16 passengers being the maximum that we’re dealing with. We are not bandwidth-constrained…we don’t have latency issues…so we haven’t had to throttle back or put prioritization. Whether it’s Ka, whether it’s using a Starlink phased-array antenna, they’re all keeping up with the demands of those passenger loads in the business jet segment. What we really did is solve for content management and in-flight experience, because we’re no longer constrained by bandwidth issues. So that has opened up all kinds of opportunities for our owners.

Claudio D’Amico

There’s an increasing amount of capacity, and there’s a step change in terms of performance being delivered to connectivity customers today. We’re addressing that with high-capacity satellites that are flexible. Our strategy continues to be to deliver these high-capacity satellites and augment our network with other types of solutions.

The idea is that we can stay ahead of the demand, especially as the network and the number of customers increase. We think we can do that with the roadmap that we have in place, even though, as the network matures and you add a lot of passengers and a lot of customers, it’s harder to sustain a certain level of experience in areas where there’s a lot of demand. That’s why we’re focusing on our GEO-backed roadmap, augmenting it with a multi-orbit strategy.

On Customer Expectations 

Claudio D’Amico

Customers have mentioned they liked the consistency and the reliability as our network grew. There are certain flight routes that we struggled a little bit with for some of the large-cabin market. In the summer, we deployed 5X the capacity over that, and the performance has improved. They have recognized the improvements, and I think that was the first step in making that step change for us. As we further integrate Viasat and Inmarsat, we can bring a lot more to our customers. And as we evolve equipment, in collaboration with our partners, we can also bring even more capacity.

I think customers also like the support. We’ve worked with our OEM partners, with our service providers, to deliver a white-glove type of support to our customers. They appreciate having that type of attention, and we’re going to continue to do that.

Jay Heublein

My issue is I can’t move fast enough to [upgrade]. To be clear, we’re not talking about Honeywell systems. We’re talking about older ATG-type technology, where the experience is nowhere near the same as that is available now. It’s a race for us to get them on. We started with all of our large-cabin aircraft. We got through those fairly quickly. Next week, we’ll start our last STC, which is for our light jets.

We think we can’t move fast enough for them. That’s the complaint we get when they get an airplane that doesn’t have the newer technology, connectivity solutions. It’s coming.

Our experience [with Starlink] has been spectacular. We’ve got thousands of hours of flight time. I’m aware of three total mechanical issues ever with the equipment. These guys know, Ka was a game-changer in terms of in-flight connectivity, but that’s relatively big equipment. [Starlink] let that experience go to all different sizes of aircraft.

Jason Wissink

Talking about Ka, the existing JetWave customers have certainly begun to experience the increased capacity over North America, especially what Claudio was referencing. We see it in the data. We track all kinds of metrics. We’re seeing peak speeds increase quite substantially and usage going up.

The reception on our newer system that we’re just finishing up—JetWave X, we bring that into the service at the end of this year—people appreciate the flexibility we built into the system. They understand what they can expect the experience to be based on, being able to access Viasat’s latest and greatest satellites.

The feedback we get is very similar to what Jay hears: “I like what I’m seeing in this lab or test flight. I want it on my airplane right now.” That’s where we work, as closely as we can with the installers and OEMs and everyone, because it takes everyone to get the certification done, get the testing done, and the install design. The faster we can do that, the faster we can truly bring these experiences to people that they’re asking for.

[JetWave X] over-the-air testing has begun, and we’ve got it on one of our test aircraft. We’ll be flying that here very shortly. Things are starting to move here pretty quickly. And now we’re getting into much deeper discussions with the service centers on STCs and production fit.

On Connectivity Improvements and Costs

Jay Heublein

Our large-cabin data subscription rates are probably half what they were five years ago. That’s because there are multiple solutions and not just because we have a LEO solution. But there has been a reduction in data costs.

Claudio D’Amico

I think there has been. It’s about which service plan you’re selecting and how you want to operate the aircraft. One of the things that we’re focused on is how do we measure that experience, so we can quantify that experience going forward. There’s a lot of talk about speeds and how you define that experience side of things. We see speed tests on our network of over 150 megabits per second. We don’t talk about that because we want to make sure that we’re consistently delivering that experience.

Jason Wissink

An interesting challenge that owners and operators have today is the challenge of choice. Choice is a good thing. When you bring new entrants into a market competition, the customer typically benefits. I think you’re seeing that in in-flight connectivity, but because in the world of equipping aircraft, switching costs are really high. It’s not like choosing between AT&T and Verizon. It’s like, I have to put the airplane down. I have to do wiring, all that stuff. So making the wrong choice can be a struggle.

People are asking a lot of questions today just to try to understand, “I get what I could put on my airplane right now. What do you think it looks like in three years…in five years?” Those are the questions that we’re fielding from customers and what we’re trying to be cognizant of as we design next-generation products.

Flexibility, integration, and simplicity are the things that we think are going to be important to give people a great experience. Because, I think Jay put it well, it doesn’t have to be complicated. People just want to get on the airplane, and they want to work, and they want to do what they want to do. Taking that as a requirement, and then figuring out how to design equipment that will meet that, those are the things that we’re trying to take into account as we work on our next generations of equipment. We’re trying to take as many of those customer-use cases and concerns into account and design something that can address some of these newer questions.

On Software Upgrades and Future-proofing

Michael Schrage

This is an architecture question. Here I am, calling in on an Apple [device], and now and then my OS gets upgraded. Here you are, Jason, you’re looking two years down the road. To what extent are you building equipment with firmware capabilities, upgradable capabilities, so that you use software enhancement as a substitute for hardware improvement?

Jason Wissink

If you look at the world of RF and satellites, there are certain frequency ranges; Ku and Ka are the main ones we use for in-flight connectivity. If you have an antenna that can serve the whole frequency range of Ka or Ku—in some antennas now it’s both—you have an antenna that can theoretically talk to any satellite. Then it comes down to do you have the right modems and waveforms, and things that you need to communicate with a certain satellite. That was very hardware-based and proprietary in the past. We are seeing the industry move towards more software-defined waveforms.

I do think there is a future where we get to much more of that, where if there’s a new network, or a new satellite, but you need to change the waveform or something, it can be much more software-defined, versus rip all this stuff off the airplane and put new stuff in. The industry is going that way, but there remains work to be done, such as standardization, to truly get to that point where it’s completely flexible.

Jay Heublein

One of the biggest challenges we went through with our partner early on was that they were constantly evolving the software, and we had to build a rig so that we could test all this in a lab and not have to deal with rollouts into the aircraft. It was a little challenging, but it’s very heavily driven by software. The technology evolves weekly, and [so does] my software.

Claudio D’Amico

One of the things that I think is important is how you build flexibility into the design of that satellite. I think there are software-defined protocols on satellites as well. You have to be able to incorporate that into your design so you can continue to serve the market, evolve the offering over the years, as technology and equipment evolve, in terms of your users, your antennas on the aircraft.

Some of the newer satellites that Viasat is launching now, we’ve incorporated a lot of those features, so that when you upgrade or you have access to that satellite, that’s sustainable over the years. It’s definitely an important factor for Viasat, and we look at how sustainable this asset is.

You have to have the flexibility to continue to evolve your service offering, to address the market over the years, and so the technology needs to match it. It’s how we think about the design.

The other thing that’s super important that we think about is security. What are the security requirements that you have and that your customers have? As you think about the design of that satellite, how secure is it? Who is using it? How can you manage that? All these are things that we incorporate as part of the asset development and the satellite development.

We have specific requirements on how certain terminals need to be able to operate on Viasat’s network. We collaborate with Honeywell, and there’s a terminal that has been very successful in this market, JetWave, which is on the majority of the large-cabin jets. Honeywell has been talking a lot about the development of JetWave X, which brings a lot of that flexibility to the market. We do that with every partner we have, so we’re in lockstep in terms of the technology, in how we can evolve the overall service offering.

Jason Wissink 

Claudio said it pretty well. We exchange requirements and all the things that you would typically do when you’re building equipment for a network or an aircraft, but there is an extensive amount of engineering coordination between the companies. We’re in each other’s labs. We’re on test airplanes together. We submit documentation to regulatory officials together. It’s two separate companies, but you almost need to function as one team. At the end of the day, the customer doesn’t care about the antenna. I wish they did, but it’s just a thing that provides an experience. We want to make sure that our piece of the value chain does exactly what it needs to do.

On the Intelligence Layer

Michael Schrage

My experience working with large organizations, not planes, is that at a certain point when you’​​​​​​ve got a lot of data streams and workflows going on, there needs to be some intelligence layer and coordination, and metadata matters as much as capacity, if not more than capacity.

I don’t believe the future of greater bandwidth and greater capacity is going to be less intelligence in flight. I think that, Jason, there will be GPUs on planes, coordinating these kinds of things, because it’s exactly what you said, Jay, people want their terrestrial experience in space. Every place where people are compute-intensive, there’s intelligence coordination going on with that.

Jason Wissink

I completely agree with what you’re saying. When we look at the history of how things typically evolved with in-flight connectivity, the newer networks come out, they always have tons of capacity. Much more than the previous generation. And performance is a step change, just like we’ve seen three or four times now.

Eventually, demand will start to outpace the network, because it’s just the nature of how things have worked in connectivity forever, and then you do have to get smarter about how you’re using the bandwidth to continue to deliver the experience that the customer is expecting. So in my view, you absolutely need both. Capacity is the baseline. But in the world of satellites, “more more more” isn’t always feasible. You do need to smartly manage the capacity, because there is a lot of background activity that happens with connected devices that people don’t know is happening, and it’s providing them no value. As you start to approach that point where things are starting to get a bit constrained, those smarts start to become very important.

Claudio D’Amico

We think a lot about that. I agree with what Jason has said that a lot of these things should be transparent to the end user. The end user is being bombarded with a lot of this information when the only thing they’re looking for is, “How is this system performing? Is this system enabling me to do whatever I want to do?” I think that’s how we have to evolve. The way we’re thinking about that is, how do we change that conversation to something more meaningful in terms of experience?

In the past, [we were] trying to train people to associate value in performance with a specific speed test. That is not a good metric. We’re thinking about four metrics. We think about terminal availability. How can you monitor your equipment? How can you get maintenance logs? How can you get real-time telemetry? We look at signal strength. That’s also part of reliability. What’s your forward link quality? What’s your return link quality? We look at performance. What’s the resilience of that transmission, measuring the network capacity to minimize packet loss? Because that’s an important one for experience. For us to do what we’re doing here on a video call, if you’re losing packets, it’s not going to be a good experience. So you have to be able to measure that. Demand satisfaction.

We’re looking at all of this from a network standpoint, and we’re trying to come up with…a metric that makes that simpler to users. I think the cell phone industry has done that when you look at the little bars on your phone. If you know you have 5G and you have five bars, you know you can do pretty much whatever you want.

How do we develop that metric, and then how do we ensure that our customers associate the experience with that metric? In the background, we can use all of these KPIs and work with integrators and with equipment partners on the network to make adjustments necessary to deliver the quality of experience that our customers are expecting. That’s the focus that we have.

Jay Heublein

We’re a technology integrator. We’re not launching the satellite or putting the constellations into space. We’re not building the core technology. We’re integrating it, and then we’re consuming it. Our decision-making was whether we could reproduce a terrestrial-like experience. That was the starting point. This is a very capital-intensive world to play in. There aren’t that many groups that can do that. We spent a lot of time evaluating financial viability and what it was going to cost to carry out the mission that we wanted to benefit from. Could we have a reliable system? Could we put standardized equipment on every single one of our aircraft? Ka is great, but I can’t put it on half of my airplanes because it doesn’t scale to a light jet or to an entry-level, midsize jet. It might at some point, but those are the things that I was evaluating.

Just to make you laugh for a second. We had a customer call in…and he went out of his way to map the data transmission speeds. He said, “I only average 230 megabits per second. I want to see 300.” And we had to have a conversation: “Were you limited in anything you or your family wanted to do?” And he said, “Absolutely not.” We created monsters looking at the data.

On the Future of In-Flight Connectivity

Jason Wissink

There will be more networks. I think the current providers will continue to expand capacity. Jay mentioned a few times, business aviation isn’t a huge industry; it’s relatively small compared to the airlines or even defense, but it’s a very demanding industry. The industry has already consolidated pretty significantly. There might be a little bit to go, but we’re getting down to the two or three that are going to serve the market. For the remaining folks who are in it for the long run, I think there is enough business.

This is a market that we remain extremely passionate about. Satellite connectivity continues to be an area that we’ll continue to invest heavily in, and our goal is to make sure we’re keeping up with the rate of change.

Jay Heublein

We’re talking a lot about consumer experience with connectivity right now, but safety-sensitive connectivity [is important], and [also] those types of things that are going to drive the future of avionics. Why do I have to have two platforms to do that? When you think about things that are going to evolve, I think that’s going to see more and more evolution [to a single connectivity system for the whole airplane]. I think eventually we’d like to get away from [separate systems]. Reliable connectivity is reliable connectivity, but there’s a lot of room to continue to evolve different aspects of it.

These are very expensive assets. So anytime you make an equipment upgrade, you’ve got the capital cost of the new equipment, but you’ve also got the cost of the downtime of the aircraft. So, a large-cabin airplane that’s $45 to $65 million, you don’t want that sitting on the ground for a couple of weeks. Even a simple system like we’re installing is still, on a large-cabin jet, about 500 labor hours. You don’t want to be making these decisions every year. That’s the big decision-making consumers have to make: “Am I future-proofing the equipment as best as possible? Am I limiting the number of times that I have to go through this experience? Am I going to get the in-flight experience that I’m expecting?”

Claudio D’Amico

What customers have liked about what we’ve offered was initially the consistency, the reliability of the equipment, and the performance. We’re going to continue to expand that. One thing that is going to be a key differentiator in this market, to be able to serve the requirements of our customers, is the support side. There’s an expectation for heightened support. You have those C-level folks or those very high-net-worth individuals flying in these jets, and when something happens—and things do happen—what is the level of support?

It’s a combination. You will have a market that is growing, but at the same time consolidating in terms of who’s delivering service. Success will come from those who are close to the ecosystem, who are developing solutions that meet the customer’s expectations in a simple, frictionless way, and at the same time, you have the whole ecosystem behind it to support it. Whoever can deliver on all those things at the same time will have a place here.

Michael Schrage 

I just want to offer [a few] insights that I’m coming away with from this conversation. The first is the indifference curve. What I’m hearing is that the indifference curve between software versus hardware is going to become more important. If Jay can buy another year of delay by doing firmware or software upgrades with the equipment, that’s probably worth a lot to him in terms of scheduling maintenance, replacement, et cetera. One of the outstanding questions is, where is software not just a good substitute but a better substitute than the core hardware in this regard? 

In terms of customer experience, I see two things going on related directly to terrestrial versus up in the air, and that is the whole notion of a handoff for onboarding. In the same way the flight attendant gets your meal in advance, they are going to get your bandwidth menu or your compute menu in advance, or there will be a button that you hit on your iPhone or your Android, and it will send a signal to the router, to the intelligence unit in the plane. And that is going to do a more intelligent job of bandwidth allocation, capacity allocation. I don’t think you’re going to have a bandwidth [split] between the front of the plane and the back of the plane. I think more intelligent allocation of resources is going to be a really interesting challenge and opportunity.