How Facebook and 500 other companies will bring 3.8 billion more people online

About 3.8 billion people in the world, or roughly half the population, currently lack internet access. The Telecom Infra Project (TIP), started by Facebook in February 2016, wants to bring the internet to all those who have yet to come online.

Connectivity gives people a voice and helps them find and share knowledge, strengthen economies, and improve communities, according to Facebook. But after bringing the internet to people in big cities around the world through wired and wireless connections, finishing the rest of the job is a lot harder. In the past year, global connectivity improved 8.3 percent, but at that pace it will be a long time before full internet inclusion is achieved.

Progress is frustratingly slow, but more than 500 companies have joined Facebook in this vast effort. The most significant expenses associated with a network are the ongoing costs of operations and maintenance. The Core and Management project groups are simplifying the core network architecture and improving efficiency and flexibility, and technologies from cloud computing to artificial intelligence are part of the solution.

All parts of the internet, from access devices such as smartphones to backhaul equipment to handle the loads of entire countries, have to be improved. The TIP group is delving into the technologies that have to be developed in the network to improve access, core and management, and backhaul. Groups such as Open Cellular, chaired by Facebook, are working to bring down the cost of telecom equipment so that internet access can become more affordable, particularly in hard-to-reach areas of the world.

I spoke with Axel Clauberg, chair of Telecom Infra Product and a vice president at Deutsche Telekom, and Aaron Bernstein, director of connectivity ecosystem programs at Facebook, about progress the group is making ahead of the TIP Summit 2018 event coming up in London on October 16-17.

Here’s an edited transcript of our conversation.

Axel Clauberg: In my day job, I’m within Deutsche Telekom’s strategy and technology innovation team. I’m responsible for our internet end-to-end architecture, as well as what we call the infrastructure cloud architecture that’s putting our service production into private clouds we’re building deep into our networks.

I got involved with TIP in the second half of 2015, when we had the first discussions with Facebook about how we can accelerate innovation for the telecom industry. We were working together as part of the Open Compute Project on data center topics. We started a discussion on how we could extend the overall telco infrastructure. I’ve been on the board of TIP since the very beginning, in February 2016. I took over as chairman when Alex Choi left in August 2017.

Aaron Bernstein: I’m very focused on Facebook’s connectivity efforts. A key component of that is the work we’re doing with TIP. We’ve been involved since the beginning, working with Axel and about 30 members of the telecom ecosystem. We launched TIP together in February 2016 at the Mobile World Congress. Now we’re proud to say there are more than 500 members of the TIP community. We’re excited about the growth and momentum we’re seeing.

VentureBeat: What are the accomplishments of TIP so far? What are the hardest problems still ahead?

Clauberg: Let me start with the hardest problem. When we look at the telco industry, we’re facing exponential traffic growth across the board. That’s driven on one side by video, but also we see augmented reality and virtual reality driving up internet traffic.

In the past Moore’s law helped us, with all the developments happening on the hardware side. We were able to keep production costs under control through investment in new hardware offering higher performance at the same price. But with that exponentially growing traffic, our production costs are increasing, while in multiple markets the ARPPU, the revenue per user, is flat or shrinking. That’s not a good combination.

What we put against that from the TIP point of view is a new collaborative effort to speed up innovation for this industry. One of the early successes of TIP was work we did in our millimeter wave project group. That describes technology where, as the name implies, the wavelength is in the millimeter range and the frequencies are in a very interesting part of the spectrum. It’s an unlicensed part of the spectrum. Normally, when we’re building mobile networks, we have to buy the spectrum before we can offer services to our users. Here, within millimeter wave, there’s a large part of spectrum in the 60GHz range which is unlicensed.

Facebook started to develop some technology in this space and produced that under the name Terragraph. They brought that work into TIP and we had a very successful collaboration between Deutsche Telekom and Facebook in this space, leading the project group within TIP. That’s one of the key technologies for us to deliver broadband services to a large group of users in an area we don’t have enough fiber rolled out yet.

Besides that, a couple of other project groups within TIP, whether it’s the open radio network area — Vodafone and Telefonica are pushing the development using off-the-shelf hardware to build to build radio access networks in a completely different way. All of these are examples of the move away from black box solutions we’ve seen in the market in the past, where the vendors have delivered products where you see a combination of hardware and software. We’re going toward a dis-aggregation. That’s one of the ways forward. We’re separating hardware and software. We can use the best of breed on the hardware side and likewise on the software side. This is helping accelerate innovation for this industry.

VentureBeat: To back up a little bit, where do these technologies apply? Is this at the cell tower, at the last mile to the consumer? How are they going to help?

Clauberg: The millimeter wave example is from the last mile. But we have other examples, like the open optical packet transport, which is right at the core of the network. In the service provider world we’re looking at the internet core, where we have very high end routers transporting internet packets with transmission speeds of 100 gigabits and beyond. We typically talk about the aggregation network, where from the core we have the link toward the access space, and then the access. But we need innovation across the board, in all of those domains, and we need innovation in the way we’re managing and operating our networks.

Automation is key for us as operators, to keep our production costs under control. It’s not just the various domains in the network. It’s also the question of how we’re managing our networks moving forward.

VentureBeat: Are some of these technologies what you might call dual use? Where they make the network better for everybody, versus making the network more accessible to people who have inferior access or don’t have access yet?

Clauberg: That’s correct. That’s the requirement from the developed market. Deutsche Telekom is an operator active mostly in developed markets. The requirements from operators in emerging markets come together with ours. In emerging markets, the issue is around connecting more people to the internet, while in developed markets it’s producing a higher-bandwidth internet service at a lower price. The requirements in the space are very much the same.

VentureBeat: As far as progress goes, we’re at something like 51 percent of the world’s population with access to the internet now?

Bernstein: I think the latest we’ve discovered in our State of Connectivity report is 3.8 billion people who don’t have internet access, which is down from 4 billion in last year’s report. I believe that’s around 51 percent, yes.

VentureBeat: If you look at the rate of change, what do you notice there? Is it happening too slowly for your goals? Do you feel like it’s happening at a good rate?

Bernstein: We actually want to do everything we possibly can to accelerate the rate of people connecting to the internet. That’s why Facebook is so heavily focused on connectivity. But to your point earlier, part of that is how we can work with operators and other partners who are focused on connecting the unconnected. Simultaneously, there’s the increasing challenge that Axel raises, in more development markets, to ensure that the amount of capacity operators are looking to carry can be handled as efficiently as possible, considering all of the parameters in play.

TIP is simultaneously focused on driving acceleration of connectivity in emerging markets, and also very much focused on helping operators and other partners in even highly developed markets improve the health of their business, so they can continue to invest rigorously in future network builds. The other point Axel makes is also important. This is not about addressing just one point in the network.

From the beginning with TIP, we established three broad tracks, what we call project groups. Those are access, backhaul, and core management. Those three tracks cover the major areas of network components where we feel like there are opportunities to drive innovation, more efficiencies, and approaches like dis-aggregation and enabling new players and new approaches to be brought forward. We feel like that plays an important role in addressing the biggest challenges that operators are facing going forward, in emerging markets as well as developed markets.

VentureBeat: This is not exactly your territory in technology, but what do you think has to happen on the social side to help accelerate adoption? There are some obvious things in some areas, like electrification, basic quality of life improvements. Is there coordination you can participate in to help this happen faster?

Bernstein: Absolutely. We also have efforts and projects with groups like Power and Connectivity, a subgroup that’s working on how to figure out making these solutions better able to perform in areas where there’s low power, or where the infrastructure may not be as stable. There are aspects of many of the project groups where the technology can account for some of those different and varying conditions.

Clauberg: There’s one key social aspect inside of TIP, and that’s the collaboration we’ve initiated here among the various players in the industry. We have members that are more traditional vendors in the space. We have startups. We have systems integrators. We have the telcos themselves. We have formed a community with people driving the internet, like Facebook for example. That’s a new way of working together in this industry, compared to the old way of innovating inside the vendor or inside the telco. It’s important to accelerate the pace of innovation.

VentureBeat: If these are some more near-term things to worry about or fix on the technology side, what are the long-term things that have to happen? With things like 5G, some people are quite concerned that the bandwidth will just be taken up by developed countries. The improvement you get in bandwidth from 5G could cause more behavior that pushes up demand for that bandwidth.

Clauberg: I agree that the demand of bandwidth will increase with the applications that become available. During my career, before I joined Deutsche Telekom seven years ago, I was working at Cisco, where I was sales CTO for the emerging markets theater. You have to keep in mind that this technology introduction is a chance for emerging countries to leapfrog and actually pass a lot of developed countries as part of that development. It’s an interesting phenomenon in the market.

With 5G, it’s not just sheer bandwidth. We should keep that in mind. Bandwidth is one part of 5G, but also, when we look at the ability to slice the network, and on one side offer a broadband slice while on the other side you offer a slice with ultra-low latency — for example, you’d use that to support autonomous driving. On yet another side you can offer a slice for a massive number of IOT devices with rather low bandwidth per device. That’s one of the key innovations coming to the table with 5G.

It’s also a great example where you need new ways of running the network and automating the network. If you extend that concept and bring in more and more slices, you have to do that automatically, because just doing manual management like you do today won’t scale. That’s one of the focus areas for us as well. We have a project group looking at end-to-end network slicing. We have project groups looking at the introduction of machine learning and AI to manage networks and automatically fix problems that come up in the network operation.

VentureBeat: Can you explain more about why expanding and improving connectivity is such a difficult technology problem?

Clauberg: The fun part is that customers typically don’t see the amount of technology which is required to produce the fixed network service. Typically our customers don’t look at it with a technology view. Nevertheless, it wouldn’t work without the technology driving all of this. But customers aren’t looking at what’s behind their router or into the telecom network. They’re looking at applications, at new services they can access via the network. That’s the focus area for our customers.

Bernstein: The geographic and regional diversity is staggering. The different types of networks, the different types of bottlenecks to connectivity that show up, are very unique sometimes, market by market. There are quite a few common concepts that come up, but fundamentally what drove us to approach this with a community-based approach is that there is no single way to improve connectivity globally. It requires digging in and understanding the bottlenecks that, in some cases, are specific to countries or regions or operators. We’re jumping in with a number of stakeholders who care to make things much better.

Understanding the challenges of an operator that’s approaching upgrading to 5G is obviously quite different from the challenges of an operator that’s struggling to have a profitable business model when connecting the unconnected rural parts of the world. What’s wonderful about TIP is it’s an opportunity to understand, to explore, and ultimately to come up with unique and creative solutions together, solutions that the ecosystem can define, build, and deploy. The community-based approach, considering so much diversity and complexity in driving connectivity globally, is the best approach.

Clauberg: One example is network planning. A lot of the planning tasks can be automated, and that’s an important piece of delivering quality to our users. In London this October we’re running the third TIP Summit, where we’re bringing the whole TIP community together. We’ll be showing what you can optimize in planning the network, optimizing the rollout and the experience for the customer.

VentureBeat: What are some of the other goals for the TIP Summit?

Clauberg: For us it’s important to get the community together. It’s easy to say that we’re driving a collaborative effort, but collaborative effort is dependent on people, and those people need to see each other. We’re also getting more members on board, which is an important side effect. We’ll have sessions open to everyone, TIP members and non-members alike, to spread the message of what we’re doing and what we’re planning to do. But we also have a set of closed sessions for the project group members to exchange their plans and drive their execution forward.

VentureBeat: This is a very big group. How do you manage everyone to get things done?

Clauberg: Most of the communication between the physical meetings is done through collaboration tools, whether it’s video conferencing or audio conferencing. We’re using Facebook’s workplace tools to a great extent for exchanging information. That’s been very valuable for stimulating collaboration within TIP. We have more than 500 members, and they’re spread across the globe. The only problem we can’t really solve is working across time zones. But the collaboration works across the globe, and that’s very remarkable.

Bernstein: The project group model is also important. Each group has one or two co-chairs that are driving each of those specific technologies, and those are where a lot of the initiatives and movement happen within TIP — the project groups, the chairs, and the companies involved.

VentureBeat: Would you liken this to something like open source development, or some other model you can follow?

Clauberg: Open source is an important element, but it’s not everything. If you look at telco networks, at mobile networks for example, this is a world where you see a great amount of patents and IP our members have developed over the years. If you want to drive innovation while ignoring all of that, you’ll fail.

For us it’s very important that a project group can decide, at the point where we’re establishing them, which IP rules they’re following — whether they’re using royalty-free models, the open source approach, or whether it would be under reasonable and non-discriminatory terms, the RAND model. Most of the project groups are following the RAND model. This is key and critical, for example, for startups bringing the ideas they’ve developed into the community. It wouldn’t be possible if they had to give everything away.

VentureBeat: How would you sum up the effort here, and why it’s important?

Clauberg: In order to meet customer expectations for the future, in a time where traffic is growing and we see new applications coming — for example, on the video side, we just made the step toward UltraHD 4K videos. 8K video is about to be introduced in the next year. We see the first TV sets already launching with 8K resolution. We see AR and VR coming. The industry needs to meet the customer’s expectations at the price point the customer expects. In order to do that, we need to continue driving innovation and increasing the pace of innovation. That’s what we’re doing within TIP.

Bernstein: We also feel, beyond the challenges of the developed market, we absolutely want to help accelerate the rate of connectivity growth in markets where there is no connectivity, or people are highly underserved from a connectivity perspective. We feel like the community embracing both what Axel was talking about, as well as those driving connectivity in emerging markets — we feel like both of those are important missions within TIP.