AT&T talks cities: “We wouldn’t do it if it didn’t make sense, but it’s not about money”
The craziness of the conference season, which has seen Enterprise IoT Insights cover both MWC 2019 and Hannover Messe 2019 in recent weeks, means content falls through the gaps. Sometimes, it is really good content, too.
One of the better conversations from MWC 2019, which has so far gone undocumented, was with AT&T’s Mike Zeto, general manager for its smart cities business. Here, we re-run the discussion, in full.
AT&T led the charge at the Barcelona showcase for US carriers, as the only one with real presence. It also led the charge, on European soil, for the operator community at large, as one of few with 5G deployments under its belt. Its ‘business’ booth looked serious-minded, like it meant business – located, as always, away from the bunfight in the main halls, just next to the main keynote rooms, where the industry sets its vision each year.
Zeto reflected on the show’s 5G buzz. “It’s crazy. Everyone’s talking about 5G. We have made really good progress. We have it rolled out in cities and we have more coming. By the mid 2020, we’ll have it rolled out, right across the US. So it’s looking good,” he said, as an opening gambit in our discussion.
Except, the US is in the driving seat with 5G. Ericsson, with an urgent interest in getting new gear depoyed, had just come off stage, on the Monday morning, to warn Europe to pull its socks up. It was a warning for the polical establishment, really, as the region deals with regulatory disparity and market fragementation.
“It’s a process at the end of the day. It’s not easy to do it. It’s a process. That’s why we’re proud of it. Because it’s hard,” reasoned Zeto.
Our discussion looked to consider the state of the smart cities sector through AT&T’s key work in cities like San Diego and Atlanta, and the new prism of its 5G rollouts. It took in use cases, business models, edge analytics, money making, and the mission for better living.
Is AT&T the best in the smart cities business, we asked. Zeto didn’t flinch (see below, and above). But we started with lighting controls in Las Vegas, which was fresh off the block at the time, and remains a key deployment for the US carrier.
Here’s the transcript, in full. As a footnote, watch this space for more content from the cutting room floor, from both MWC 2019 and Hannover Messe 2019, in the next weeks.
Can we start with the Vegas announcement(see here for details of the Las Vegas deal)? What is the significance of that work? What does it tell us about the smart cities space?
So, Las Vegas deployed intelligent sensors for lighting controls, and some edge processing units. The significance is the city can add aftermarket sensors, like air quality sensors, as it goes
Is this new? Can sensors not be retro-fitted to other smart lighting setups?
No, they can be, if they are cellular connected. But what the Ubiquia unit allows is to manage the lighting controls, provide public Wi-Fi, and then access an edge processor that you can plug whatever air quality sensor you want into – or whatever camera you want into.
Because there’s a lot of infrastructure out there already today, and this allows for a modular approach to building out an IoT network for the city. That’s why it was important.
So that sounds key. Because that is the complaint about the broad IoT ecosystem, that it’s fragmented, and it’s hard to patch this stuff together.
Yes. So that’s a big part of it for cities. IoT is a challenge for anybody because no one company does it all. And the approach we’ve taken at AT&T is to move beyond just connectivity – to provide some end-to-end solutions that make it easier for both the public sector, which will be a third of the IoT revenues across the world, and the enterprises to have end-to-end solutions packaged up and ready to go.
We are trying to pull the ecosystem together for them, to make it easier to drive value. The work with Ubiquia is an example of how we’ve bundled connectivity with the application, and used our relationship with the city to discuss more than just lighting controls – to talk aboyt public safety, environmental data, and citizen engagement. And sort of bundle all of that up.
To be clear, because my touchstone for the AT&T smart cities programme is San Diego, which seemed quite a seminal piece of work (see here for details of the San Diego deal), which is still developing. Is that not the same as this new work in Vegas?
It’s similar, and it’s almost a category for us; so think of ‘digital infrastructure’ as a category. One of the offerings we have, which is AT&T Digital Infrastructure, is what’s deployed in San Diego. The same solution is also deployed in Portland and Atlanta, in partnership with Georgia Power. This is a modular approach to that.
So does this mean you can build more easily on top than you can in these other cases?
Well, the setup in San Diego has two cameras, an environmental sensor, and an audio sensor all built in – so more capabilities right out of the box, versus a modular approach that allows you to add sensors and applications as you go. It gives the customer choice, if they need that in their strategy.
And, to be clear, does the San Diego setup allow for the addition of new sensors and applications, or does it preclude such a modular approach?
You could attach something after-market; there’s a port to attach after-market. It’s a little bit modular, if you like. It’s like a loaded car, versus a car that has a great package, which allows you to choose what you want. In cities, they are approaching it in different ways. Some want to start with LED lighting, and put controls on. Some want to provide public Wi-Fi, and other apps and services from the start.
Having both offerings is really beneficial to the city. And in some cases we’ve responded to bids with both – so the same AT&T Digital Infrastructure that’s deployed in San Diego, and then lighting controls and public Wi-Fi on top.
I guess the cities that went early on this had it plotted out already, and had the maths in hand, whereas others are interested, but more risk averse.
Yes, it’s less expensive, and you can increase the investment as you see success. So San Diego procured 14,000 LEDs with controls and 3,200 sensor nodes to begin with. It was a $30 million investment upfront. You go the other route, and you just deploy LED lighting and lighting controls, and then deploy the additional modules where you need it. It’s a less expensive approach to the same problem.
Can you talk about the Vegas investment?
I can’t say what the number was because it’s pilot, and not public. It’s a smaller pilot – 60 sensors were deployed in one quadrant of the city, in its innovation zone. So it’s a much smaller investment, although it’s hopeful itwill expand. The city is very pleased with it, and a lot of applications can be leveraged to drive value.
Was that on the back of a 5G discussion? It seems the discussion about 5G in cities is opening up this smart cities game, and you can come in on the back of it.
That was actually LTE-M. And it could equally be NB-IoT, coming up. For lighting controls, you don’t need a lot of bandwidth. So it’s a perfect use case for LTE-M and NB-IoT. Now, if you add cameras, then it’s a different story. Then you want broadband capabilities; you want 5G capabilities right from the edge unit.
Because it gives you the ability to take the links out of the network, to leverage edge compute and AI right there, and solve problems like curb management and traffic. There are a lot of different ways you can leverage video to drive value.
Can we talk about the edge piece, and what’s happening in the processing unit? Presumably everything is filtered out, except the important stuff – so it only signals the control room when there’s a punch up or a gun shot?
Yes, and Portland is a pretty good example. Portland is using AT&T Digital Infrastructure to drive its Vision Zero strategy, to get road fatalities down to zero. Portland’s not taking any personally identifiable information from the sensors. It is just trying to determine what it is – whether it’s a pedestrian, a cyclist, a scooter, and whether’s it’s a commercial vehicle or a private vehicle – and where it is heading, in terms of direction and speed.
And the city uses this data to determine where to put additional infrastructure – street lights, crosswalks. It’s processing all of that right there, at the edge, and only sending the metadata back. It’s hard to say how much data is being processed because it depends on the time of day, how many people are using the streets.
Is some being uploaded to the cloud?
The metadata is being uploaded, and you can set how long you want to keep the data at the edge.
Are these kinds of business cases still being made around the LED savings (see here for the business case for smart LED lights)? Is that still the way into this?
The savings from LEDs and lighting modifications are certainly driving a lot of value. You can use the savings to reinvest in emerging technologies like digital infrastructure. That’s a clear-cut ROI and business case. Now, where we need proof-points, and why we’re doing some smaller engagements of 200 and 250 nodes, we are looking to test out those other use cases. Can they get enough data to get the information to move crosswalks to the right place, and save a life.
But that’s a hard business case to work – the cost of a life, and general calculus of hard ROIs versus soft ROIs (see here for a full report on how to make smart cities pay). And it’s complicated to add up because you bring in a lot of other parts of the city, like healthcare.
Yes. An easier one is curb management. If you think about the curb map; the most valuable assets a city has are the vertical infrastructure, the light poles, and the curb-side, whether it’s for bike and scooter sharing, or for parking for Uber and Lyft and taxi cabs.
Cities used to talk about using digital infrastructure, like in San Diego, for parking enforcement and parking revenue – to get the data to know whether a spot was empty or full, abd decrease congestion. But cities are starting to pivot to the concept of curb management, and business models that can come of that – like flex-time at the curb.
And presumably to drive footfall into shops, linking with retailers?
There’s masses of ROI around that. When you look at the ability to maximise the value you can drive from the curb, it goes well beyond just ticket enforcement.
So are you guys making money out of smart cities at the moment? Because the IoT sector, at large, is not making money easily. For a carrier like you, juggling a number of technologies and solutions, how is business in this space?
We were the first carrier to stand up a dedicated smart cities business, and that was three years ago. We are doing many commercial engagements. Business is profitable. And I would say we are the leader in the space from a carrier perspective.
Definitely in the US, and we’re probably top one-or-two globally from a carrier perspective. For sure. What I can say is we wouldn’t be in this business if it didn’t make sense. But we recognise there is a long tail. This is the public sector. It can’t just be about money. There’s a lot more goodness that can be driven as a large corporate. You need that approach.
And that’s why public-private partnerships come into play. Because when you start to enter into these partnerships, you’re capturing a part of the value chain, and not all of it. The only way these cities are going to be move forward is with public-private partnerships, and business models that, frankly, might not provide a lot of short-term value for the private sector, but will do a lot of good for the community – where our employees live, where our enterprise customers live, and where we lice. There are a lot of positives.
It’s hard to break a business down, like yours, where everything is so interconnected – your standard cell deployments, your IoT stuff, your C-V2X and everything you’re doing.
I could work across 10 office groups inside of AT&T – 10 different business units. Smart cities sits in the middle of all that – construction and engineering, 5G deployments, fibre deployments, external legislative affairs, which is our local teams on the ground.
You’ve got our business solutions team, our public sector team, our IoT team, our corporate sustainability team. We sit at the intersection of all of that – of many, many business units coming together – to provide a lot of value to a city.
Is that getting easier from an internal point of view? Because I can imagine, rather like the city CIO knitting his departments together, there’s a job your side, as well, to knock heads together.
When we stood the business up, I was very fortunate to hire a great team. And within six months, the other units across AT&T saw the value we could provide if we all came together in these communities we all live in. And they’ve been very supportive. I’m blessed they’ve embraced the vision.
Can I ask about CBRS in the US, how that plays in the smart city space, and the bigger picture of spectrum liberation? Because unlicensed spectrum is coming available through 3GPP in the 5 GHz and 6 GHz bands. There’s MulteFire, too. Germany is making available the C-Band for industry. There is in the background this existential threat, almost, to operators running in licensed spectrum that suddenly anybody can go and set up a LTE network (see here and here for discussion of this dynamic, as it played out at MWC 2019 and Hannover Messe 2019, respectively).
It may be a simplistic view, and it’s not just about plugging a network in, but negotiating with vendors, and upgrading software and devices, and putting stuff through the whole cellular ecosystem, which you have been part of for so long. Can you talk about this from a cities point of view?
Here’s what I would say. We’ve invested $145 billion since 2013 in our networks in the US – that’s the investment in spectrum we’ve acquired and in networks we’ve deployed. If this business was easy, everybody would be in it. And they’re not. At the end of the day, there will always be different players that drive the market, and we’re one of them – from both an innovation perspective and a network deployment perspective.
We addressed unlicensed spectrum a while back, when there was Sigfox and others playing in the space. And we’ve done that with LTE-M and now NB-IoT. And even in the utility industry, where they tend to lean on building their own networks, we’ve been able to come to the market with a private LTE offering, leveraging the in C and E blocks in the 2.3 GHz spectrum.
So I think there will be other entrants, leveraging different types of network technology. But I feel really good about our position, and we continue to innovate. And when somebody comes up with one offering, we tend to come up with a counter.
So, to be clear on the utilities point: will AT&T make use of the CBRS band? Is that something you’re looking at?
In the utilities market, right now, we have a private LTE offering, where the utilities will lease the spectrum from us, procure the hardware from us and service it, and point towards that 2.3 GHz spectrum. So they’ll have a standards-based network, which is secure and redundant, just like they want.
And will they manage it?
They’ll manage it, yes. Just like they do today. They want the redunandancy and control, to manage the network – but they want standards-based LTE, which makes that ecosystem we’re talking about accessible.
Does it matter where they are? Out in the sticks, or on an oil rig…
The power’s low, the cycles are low. It’s all monitored and measured by the FCC and by us. It’s all tested. So there’s no interference. That’s the purpose of all the testing we go thrugh – to make sure there is no interference, that the quality’s there that they need.
Is that live now, with utilities?
We’re getting it up and running.
How are different regions doing? We’re in Barcelona, which is held up as the archetypal smart city. But it’s debatable how much is actually live, or at least scaled out.
It’s live, but, yes, I don’t know how much is scaled. But you see that across the world, where there’s a smart corridor or section of a city. Nobody has really deployed everything they could to make a city smart. I mean, you see some really large camera deployments, some large intelligent traffic systems, some large parking deployments.
But you don’t necessarily see it all together in one city. And even if they did, you’d still want all that data in one place, and cities haven’t got there.
Are you close to doing that – scaling multiple solutions citywide? Is that what San Diego and Atlanta are starting to look like, for instance?
San Diego would be one of those cities that’s on a path. Even though Atlanta is a small deployment, it’s well positioned to do that, because we’re partnering there with a utility, and that’s how we’ll be able to fund the scale-out.