Cloudflare TV

🌐 Project Pangea: Helping underserved communities expand access to the Internet for free

Presented by Jocelyn Woolbright , Annika Garbers, Marwan Fayed
Originally aired on 
Impact Week

Transcript (Beta)

Hi, everybody. My name is Jocelyn. I am the program manager on Cloudflare's public policy team, and I'm joined by Annika and Marwan on the Cloudflare team.

And we're actually going to be talking about one of our new projects that we launched today.

It's called Project Pangea and how we're helping underserved communities expand access to the Internet for free.

So I'm going to give a little bit of a background of some of the issues and talk about a little bit about the project, and then I'll introduce Annika and Marwan, and then we'll get started.

So if many of you don't know, it's actually surprising that nearly half of the world's Internet doesn't have access to the Internet.

And even if they do have access, it's either unreliable, or it's expensive, or the access is not very good.

So what a lot of communities have actually been doing is they rely on nonprofit organizations or local community groups to actually build their own networks.

So this means either erecting Wi -Fi antennas or laying fiber optic cables.

So this actually allows the group to build a network in a town, so it's connecting different houses and businesses, but it actually is really expensive to connect those communities to the Internet.

And that's because bandwidth is either really expensive or unavailable in different types of regions, and that's because different providers don't have the economies to scale this or different policies or regulations that might favor certain markets.

So this problem, one of the most surprising things I've learned is that this problem is that only limited to developing nations or rural areas, but it's actually in a lot of places around the globe.

So today we're launching Project Pangea to offer a chance for these communities to actually get Internet experience that they deserve.

So we're providing a more secure, more reliable Internet experience, and we're going to talk about how the difficulties these communities face and also how our new project is kind of an effort for Cloudflare to help these communities actually connect to the Internet.

So thanks Anika and Marwan.

Let's do like a brief round of introductions. Anika, do you want to go first?

Sure. So happy to be here. I'm Anika. I'm on the product team at Cloudflare.

I'm the product manager for some of our network services, including some of the ones that are under the umbrella of the services that we're offering to participants in Project Pangea, which we'll talk about a little bit more later.

But excited to be here and talk about the project. I'm Marwan Fayyad. I'm on the research team.

My former experience is the one that matters here, I suppose, as a professor in computer science and a co-founder of a community cooperative ISP that was launched specifically to try and solve this problem.

So it's lovely to see that Cloudflare teams take this up as a cause. Yeah, we're super excited for that.

So Marwan, I actually want to go back to you because your interesting background in community networks.

Do you mind talking a little bit about how you worked with a community network, what that space looks like, and maybe kind of talk about what a community network is?

Because it's a very specific niche part of the Internet community.

Do we want to start off there? Sure.

So I want to preface all of this by saying one of the reasons that I'm sitting here with you, and I'd argue that one of the reasons Pangea exists at all as a project, is because of the numerous people in the background, in their own regions.

They've set up community networks, many of which I've worked with, a small fraction of the world.

And in particular, a couple of other professors and network engineers with whom I co-founded this previous organization, Hubs, in the interest of transparency, to try and solve this problem.

So this started around 2010, 2011, by just trying to help communities put up their own little networks, mostly Wi-Fi at the time.

And you would think that Wi-Fi, you couldn't really do it. But if you get a directional antenna, you can buy these off the shelf for $200, $300.

They look like little satellite antennas.

And you can build really high capacity, reliable links over long distance with these things.

So last I checked in the ISP, our longest link was, I think, close to 40 kilometers or 25 miles over water from two hilltops.

And you get about, when we first put it up, it was 30, 40 megabits.

This is 10 years ago. And I think we're up in the hundreds, maybe low hundreds.

So the point here is people can put these things up fairly cheaply. What we find really, really works is when they set up, as you alluded to, Jocelyn, some sort of a not-for-profit or a community interest company.

And the reason to do that is because everybody, when they participate, you could argue that they share in the profits.

But usually what ends up happening is they take all the revenue and they put it back into the system.

So if they couldn't afford to build something the way that they wanted to initially, they can always do so later.

One of the best success stories, one of these communities that started early on the West Coast, they were all Wi-Fi and they now all have fiber to the home.

And they're in the middle of nowhere, literally.

I mean, they're 200 kilometers from anything urban.

So good on them for being able to do this. That's great. And Anika, I'm curious, when Marwan talks about community networks, does your product manager light bulb go off and being like, hey, these are real struggles that communities have.

How do we fit products to help these networks? I'm curious, whenever you first heard about the project, what was going on in your mind?

Yeah. I think ultimately product managers are problem hunters, right? We're always looking around to see what are our existing customers or potential new customers?

What are their pain points? And as Marwan was describing this space, I was so surprised over and over again by just the number of things that are in that list that ultimately add up to the problem that we're trying to help solve here, which is half the world's population doesn't have reliable access to the Internet.

I mean, some things that surprised me were just the physical, literally physical barriers in the world.

I think often at Cloudflare, we think in terms of software and stuff sort of further up in the stack, you could call it, but we're talking about in some cases, problems with animals chewing through fiber cables or hills blocking wifi signals, or literally in one of the conversations that we had about this project with a community network organizer, she was able to pull up a map and show us the actual physical distance between some of these communities that are struggling to get Internet access and the closest place with an Internet exchange point, which is somewhere that they'd be able to actually get that traffic to the rest of the Internet.

So lots of physical barriers, and then also lots of people complexity too, and just political things and different factors that make this problem a really challenging one to solve for lots of reasons.

It's interesting that you mentioned that.

I think a lot of people, when they think about the Internet, you automatically connect on your computer or your phone, and you don't realize they're actually cables in the water that are connecting us to many different regions of the world.

So kind of thinking about the community network space, so Marwan, how does it look?

So we have civil society organizations, you have tech companies, you have government agencies.

How many different pieces are there when it actually comes to connecting these community networks to the Internet?

Oh my goodness. That's a loaded question. Answering that question.

So I can tell you, it all starts at least for me, my first exposure to this, and I might be showing my age a little bit here.

This almost predates America Online.

It was sort of one of the biggest home broadband ISPs on the planet, served Europe and North America at the time.

And this is strictly dial-up, strictly speaking, so 28 kilobits per second.

And I remember reading about people who were in the farming communities, they're so far away, the copper line that serves the phone, while it was fine for voice, it was so far away that the signal was too weak and too error-prone to carry data.

So what they would do is the person on the farm who lived closest to the cabinet would actually put up antennas, and they would, and this is how they would all share their data, all the way down forest farms over tens of miles.

So when we talk about the space and how many parties are involved, on the surface you'd think it's just the community, which is okay, except communities are full of individuals who all have different incentives and different expectations and different priorities.

And so we have this vision that you just walk in and build it and people will come, everybody will want it.

It rarely works out that way. Usually it's a very small portion of the community will decide it's something they want.

Other people think, ah, I've done thus well without, I don't need it.

And then they see their neighbors who have it and the benefits, so they want it and it kind of spreads from them.

And then you think, okay, fine. Then the community's all set up on the, getting to the outside world, it's a matter of picking up the phone and arranging whatever kind of connection.

The truth is, these are really hard to find. And if they can be found at all, a place that you can connect to, no single community can afford them.

So that already by making the call, you're having to shop around for all the other players who might provide you those Internet services, only to realize you can't afford what they have to offer.

So then you get involved with your community groups and special interest groups and you form your own and you petition government, various layers.

And every one of them is enthusiastic, but they are cautious by nature.

So you're sitting on the civil service side and you think, well, I can hire a large incumbent.

Surely they should be able to solve the problem.

Only the problem has persisted now for more than two decades. And so there's this push pull with the civil service.

There's way more players in this than I think most people would start to think about.

To think about the different players, can we talk a little bit about what do you think the largest challenge is when it comes to Internet access?

And why are there so many people that have issues getting coverage?

And Monica, if you want to jump in, because I know you've talked a lot about the different issues that people have actually connecting to the Internet.

Sure, sure.

So I mean, Marwan started off by talking about the community networks and you mentioned sort of the local connectivity aspect, right?

So it's less challenging than you might think to do something like set up Wi-Fi in your community.

But then there's sort of two other pieces that you need to put in place.

The analogy that we've been sort of using is like roads in a town versus an on -ramp to the highway.

Getting the on-ramp to the highway is really challenging for a lot of these communities.

And that's the piece that Cloudflare is hoping to help with.

I wonder, Marwan, do you want to break down sort of the backhaul versus transit and the different options that communities have here?

Sure. So let's just assume that the community has their own little local network or a few people have their own network and then they want to connect to the outside world.

So oftentimes what will happen is you will find a local connection point and local is relative here.

It could be 50, 100 kilometers down the way. But there's a place where the community can put its own router and then from there a patch cable into the router of the next provider.

So we think of this as backhaul, tends to be how we refer to it.

And very often what happens is you will end up buying both the backhaul to some exchange point that is typically in the city as well as the Internet service from one party, because to bundle it is cheaper.

But it doesn't necessarily have to be that way.

So you could go to a traditional provider, typically telecommunications or energy, but maybe you can also convince an educational network or a research network, for example, just to carry your packets from close to you to somewhere in the city.

And then once you get closer to an exchange, you can buy service from anyone you want.

And this is the critical component.

So we have this notion that as soon as you connect to a large provider of some kind, the carrying the packets and the Internet service are the same and they don't have to be.

So you could buy them both together or buy one or get a backhaul from one and then buy the Internet services from a third party.

So you mentioned these two different challenges.

So Anika, how does Project Pangea and how do Cloudflare's products kind of fit in and how can we help kind of connect these community networks to the Internet?

Sure. So I think actually it might help to take a step back and explain the point of an Internet exchange in the first place.

So there's a scene that I always kind of giggle at.

And I think it's like one of the Avengers movies where they're like going to the center of the Internet.

And that's funny because the Internet by design is sort of decentralized, right?

There isn't a center of the Internet where all data flows through.

But there's some sort of truth to that in that there are what you could think of as sort of hubs.

There's places called Internet exchanges where people that carry packets around the world.

So that includes different large telecom providers, people that have really large networks that are just pushing a lot of traffic around, the Googles, Facebooks, Amazons, Microsofts of the world.

And then also pretty much anyone else that's trying to get access to the content that's served between any of those providers, anyone that's trying to sort of just get an onramp to the Internet, they can basically all come to that one place.

And then there's a building or a set of buildings where everybody's just sort of all peering with each other and exchanging traffic around.

And so Cloudflare, because of the way that we've architected our network over the past 11 plus years, is to be very close to where users are and very close to where content is so that we can serve it very quickly and protect from security threats at the edge and things like that.

And so we're already in a bunch of these places.

Cloudflare's network is in over 200 cities and over 100 countries across the globe.

And so our thought process here was, Marwan broke down the backhaul versus Internet services piece.

If a community network can figure out the backhaul portion of that, if they can get to an Internet exchange point, any of these cities that have them, because Cloudflare's in the sort of vast majority of them, then we can help by providing the Internet services piece.

So actually get the packets from that network to the Internet and back, and then also layer on some services on top of that that help keep the traffic protected, fast, secure.

So things like DDoS protection and a built -in network firewall are also included in this service for Pangea participants so that they can make sure that their networks are reliable and also protected from threats that are out there on the Internet.

I'm going to add to that a little bit, Jocelyn, and say one of the reasons this matters, this is arguably, it's the one piece of the puzzle that is completely outside of the control of the community networks.

So they can build infrastructure locally, they can even band together with other community networks that might be nearby in order to get better backhaul services to an Internet exchange.

All of these things, they have a say in, they have some ability to direct their own destiny.

But when it comes to those Internet services connecting to the rest of the Internet, this is the one piece that is completely out of their hands.

And so it's a big deal that this is on offer. Yeah, and that's, I think, where we recognize that a lot of these security products that we provide might be able to kind of fit into that piece.

And I like, Annika, that you mentioned that the Internet, there's so many different players that play an essential role in making sure the Internet stays online.

And it's just like community networks, like Marwan, you said, like civil society organizations play a huge role in helping these community networks build their network, but also connect to the Internet.

And one of the challenges that I think is probably the most interesting that I didn't necessarily recognize is that when you think about a community network, a lot of the issues are in terms of terrain and in rural areas.

But then you would see community networks, for example, like NYC Mesh, which is in New York City in a very highly populated and dense area.

So what do you think the challenges are? How do different networks, what do these challenges look like for different networks?

For example, New York is going to have different challenges than a network, for example, in Nepal or the Republic of Georgia.

So Marwan, do you want to talk a little bit about that?

Boy, so this is purely anecdotal and based on personal experience.

There's this notion that because you're far away, if you're not in a rural, then it must mean that you're low income.

And so that's the raison d'etre.

And it turns out to be actually entirely untrue. So communities everywhere are just like communities everywhere else.

And just because a community is far from a city in no way, shape or form means that they are unable to pay.

Usually they are more than able to pay. They just have no access. The urban regions flip that around just a little bit, where there's lots of access everywhere you look.

But in the grand scheme of bills that come up every month, not everyone is going to be thinking about Internet access as the single highest priority.

So the main motivation for community networks outside of cities is access.

The main motivation for community networks inside of cities tends to be a little bit more egalitarian.

And NYC Mesh very much operates in this space. So as I understand, you can connect to NYC Mesh for nothing or next to nothing.

Or you can choose to subscribe as well as a way to contribute to the project.

So it can go both ways.

Yeah, that's that's a great point.

So you all have both had many conversations with community networks that we've been chatting with about Project Pangaea.

So for example, NYC Mesh, Hubs in Scotland, Alter Mundi in Argentina, Zen Zeleni Network in South Africa.

So Anika, I actually want to start with you. So what is one of the most surprising kind of items you've learned in these conversations with them and something you weren't necessarily expecting?

And then Marwan, I want to ask you the same question, because, you know, you have this background in community networks.

But now that you've kind of talked to community networks around the world, like, what have you learned so far?

And Anika, we'll go with you.

Yeah, I think, initially, you know, starting out with this project, when Marwan kind of approached the folks that work on the network services and came up with this idea for, hey, could we offer something like this to these community networks, I knew very little about this space.

And so what's surprising to me, maybe not surprising to other folks that were a little less naive here, but I think just the, as a product manager, I think when you hear a lot of people articulate the same problem over and over again, you know that you're going down the right path, right.

And Marwan, you know, pitched us very confidently that like, this is the problem, or this is a really huge problem that's facing these community networks.

And that's what we heard over and over again from these participants, right?

Like, this is a huge part of their cost, or it's a huge barrier to being able to complete that last piece in the puzzle to grant them access.

The, you know, the fact that rural and remote communities can sometimes be, you know, just as the same definition sort of underserved in terms of the amount of access they have as the urban communities.

And then also just sort of the, there are fun stories, right, about the animals and the hills and the struggles with, you know, convincing people to go one way or another as political tides change and literally, you know, governments and entire countries have different positions on things.

It's been a really interesting project to learn a lot about them.

And I think if you're, you know, like me, someone who's starting new to this space and wants to learn, we have information about all of the folks that are signed up as participants for Pangea on our landing page.

So you can go and learn a little bit about each of their individual stories and journeys, because there's some through lines through them, but also a lot of really unique stuff in each of their own sort of individual journeys to help get their communities Internet access.

So same question for me, I guess, what have I, I tell you, there were many surprises en route, but in terms of the conversations we've just been having as part of this project and its development, and this one's a little bit difficult, I think, for me to articulate.

It's one thing, it's one thing to sort of have this background knowledge that there's a lot of activity, but it's another thing to somehow bear witness to it.

And I think what was heartening, like it was heart filling, but at the same time, also devastating was how much progress is actually being made in different parts of the world.

And I had no idea. I mean, so I've been operating in this space in one form or another for the better part of 10 years.

And still, I was surprised by how well people were doing, crucially, in regions you wouldn't expect to see it.

This is three or four places in sub -Saharan Africa, the Argentinian network, Alternwindy, my goodness, the progress they've been making just blows your mind.

In South Africa, they're thinking about building networks, at the same time that they're trying to build reliable water supply to some of these places.

I mean, that's visionary in a sense. And so very often after these conversations, I just have to turn around, shut down and take a deep breath, because it was a little bit overwhelming and spectacular to hear.

Just shows how much Internet access has become a type of dependency and how we look at it as an essential reliable good for us.

I'm actually curious about what are the technical requirements of your community network that is interested in joining Project Pangea?

Anika, do you want to talk about what the requirements are and maybe how we can help some of these community networks?

Sure, sure. So there's three pieces of the Cloudflare product umbrella that are offered under Project Pangea.

The first one is Cloudflare Network Interconnect, which is just a physical or virtual connection to be able to get traffic onto our network.

So that's requirement one, is that the community network can establish that backhaul piece from their network to ours.

And often in many cases with these networks that we've talked to, that's turned out to be relatively easy because they're already located in the same place that Cloudflare is in the building.

So it's just a matter of running a cable from one room to another or something like that.

So that's piece one.

The second piece is that in order to attract traffic to that community network through Cloudflare, we advertise a customer's or our participants IP space on their behalf.

So a requirement for the piece of this puzzle called Magic Transit is essentially bringing your IPs to Cloudflare and we advertise them to the Internet with BGP.

And that's how all the traffic, good and bad, ends up at our network.

And so participants at this stage need to have a slash 24, a C block of IP space, which is the minimum sort of accepted on the broader Internet with BGP advertisements.

And over time, we're going to look at relaxing those restrictions and supporting things like IPv6 and potentially the ability to offer these services for networks that don't own or at least borrow any of their own IPs.

So those are the primary pieces from the technical perspective.

And then what they get by doing this is all the traffic coming inbound to their network and then also out to the broader Internet comes through Cloudflare.

It's all filtered through our industry leading security tools, again, like the DDoS protection and network firewall, which are really important pieces because we've also heard stories, unfortunately, of small networks that have gotten to the point where they have everything set up, they have an upstream service from a large provider, and then they get hit by a devastating DDoS attack and that upstream service is shut down, which is really unfortunate because it doesn't make sense for that provider to sort of continue to serve their traffic if their links are getting congested because of serving that network.

And so we understand and recognize that the security piece is really important in this equation and that's something that Cloudflare has as sort of a core competency that we're offering along with that.

I think that's the one. Did I miss anything from the technical perspective, Marwan?

Yeah. Sounds about right to me. Yeah, I think it's really important to highlight the security side because we talk about accessibility for community networks, but sometimes you might forget about the security side and what types of threats these community networks might face.

So it's also good to have the performance and reliability side, but also the security side of that.

So we actually have a question from the audience from Steven.

So whoever would like to take this, the question is, how will the rise in consumer satellite Internet technologies impact underserved communities?

Okay, I'll have a go.

So Steve or Steven? Steven, sorry. So it's been said before that I'm an optimistic cynic.

So I'm going to offer the following with that hat on.

I wholly applaud a lot of the low Earth orbit satellite stuff that's happening today.

I think it's brilliant on so many reasons. I will acknowledge I was a naysayer initially, and I'm really, really impressed to see it get as far as it does get, as it has gotten so far.

I will point out a few things.

The first is the signal is still far from reliable. By the way, so satellite here has a long history of saying that it's going to solve the problem, long, and just failing to do so utterly.

So charging in the range of hundreds of units of currency and to provide service that is a fraction of what you would get if you were closer to a city.

And that's been the tradition. And I don't see that low Earth orbit satellites are going to solve that entirely, maybe for a very, very small portion of people on the planet.

But I claim community network solutions are better where available.

They're less variable. So low Earth orbit satellites right now, they go up and down and up and down there.

If they're done right, you don't pay any more than you would if you were living in a city.

Oftentimes you pay less. There was a moment when the west coast of Scotland was getting 40, 50 megabits per second at about 20 British pounds per month.

And if you lived in Edinburgh, you were lucky to get 20 megabits per second at 30 pounds per month.

So the community network solution is good, not just on a technical level, but because it actually maintains a level of health within the community itself.

Satellite can offer that. Yeah, great, Marwan.

Thanks for that. Great question, Stephen. So we have about a minute and 30 seconds left.

So I want to go to each of you and say, do you have any last minute kind of notes or things that you've learned about this project?

What you kind of see the future of Project Pangea?

Annika, I can start. I'll start with you.

Just that I'm really excited. I'm really, really excited about this and to get our participants up and running.

If you are in this space or you know someone who is, please tell them to apply.

It's slash Pangea. It's the landing page that has the full information about the eligibility requirements and then a form you can submit to give us your information.

So please reach out if you are someone that you know is involved with something like this.

From my side, I think it's important to acknowledge, by no means are we offering to solve all of the community's problems.

But when the community manages to get together and can get to the exchange, Cloudflare will be there absolutely waiting to provide the service.

And just as a tip, knock on the doors of the educational and research networks that you have nearby.

If you have universities and the other publicly funded networks on a personal level, taking off my cloud, have an obligation, I think, to fill this gap, at least in the interim until you're self-sustaining.

And let me tell you, there's lots of precedence for it and that it makes all the difference.


Thank you so much, Marwan and Anika, for joining us. This has been really great and have a great day, everybody.

Thanks, Jocelyn. Bye.

Thumbnail image for video "Impact Week"

Impact Week
Tune in for all of Cloudflare's Impact Week programming, featuring an array of CFTV episodes spanning environmental, social, and governance issues.
Watch more episodes