Categories
Regulation - What is it Good For? The Industry

Digital Discrimination

The FCC recently opened a docket, at the prompting of federal legislation, that asks for examples of digital discrimination. The docket asks folks to share stories about how they have had a hard time obtaining or keeping broadband, specifically due to issues related to zip code, income level, ethnicity, race, religion, or national origin.

The big cable companies and telcos are all going to swear they don’t discriminate against anybody for any reason, and every argument they make will be pure bosh. Big corporations, in general, favor more affluent neighborhoods over poor ones. Neighborhoods that don’t have the best broadband networks are likely going to be the same neighborhoods that don’t have grocery stores, gas stations, retail stores, restaurants, banks, hotels, and a wide variety of other kinds of infrastructure investment from big corporations. The big cable companies and telcos are profit-driven and focused on stock prices, and they make many decisions based on the expected return to the bottom line – just like other large corporations.

There is clearly discrimination by ISPs by income level. It’s going to be a lot harder to prove discrimination by ethnicity, race, religion, or national origin, although it’s likely that some stories of this will surface in this docket. But discrimination based on income is everywhere we look. There are two primary types of broadband discrimination related to income – infrastructure discrimination and price discrimination.

Infrastructure discrimination for broadband has been happening for a long time. It doesn’t take a hard look to see that telecom networks in low-income neighborhoods are not as good as those in more affluent neighborhoods. Any telecom technician or engineer can point out a dozen of differences in the quality of the infrastructure between neighborhoods.

The first conclusive evidence of this came years ago from a study that overlaid upgrades for AT&T DSL over income levels, block by block in Dallas. The study clearly showed that neighborhoods with higher incomes got the upgrades to faster DSL during the early 2000s. The differences were stark, with some neighborhoods stuck with first-generation DSL that delivered 1-2 Mbps broadband while more affluent neighborhoods had been upgraded to 20 Mbps DSL or faster.

It’s not hard to put ourselves into the mind of the local AT&T managers in Dallas who made these decisions. The local manager would have been given an annual DSL upgrade budget and would have decided where to spend it. Since there wasn’t enough budget to upgrade everywhere, the local manager would have made the upgrades in neighborhoods where faster cable company competition was taking the most DSL customers – likely the more affluent neighborhoods that could afford the more expensive cable broadband. There were probably fewer customers fleeing the more affordable DSL option in poor neighborhoods where the price was a bigger factor for consumers than broadband speeds.

These same kinds of economic decisions have been played out over and over, year after year by the big ISPs until affluent neighborhoods grew to have better broadband infrastructure than poorer neighborhoods. Consider a few of the many examples of this:

  • I’ve always noticed that there are more underground utilities in newer and more affluent neighborhoods than in older and poorer ones. This puts broadband wires safely underground and out of reach from storm damage – which over time makes a big difference in the quality of the broadband being delivered. Interestingly, the decision of where to force utilities to be underground is done by local governments, and to some degree, cities have contributed to the difference in infrastructure between affluent and low-income neighborhoods.
  • Like many people in the industry, when I go to a new place, I automatically look up at the conditions of poles. While every place is different, there is clearly a trend to have taller and less cluttered poles in more affluent parts of a city. This might be because competition brought more wires to a neighborhood, which meant more make-ready work done to upgrade poles. But I’ve spotted many cases where poles in older and poorer neighborhoods are the worst in a community.
  • It’s easy to find many places where the Dallas DSL story is being replayed with fiber deployment. ISPs of all sizes cherry-pick the neighborhoods that they perceive to have the best earnings potential when they bring fiber to a new market.

We are on the verge of having AI software that can analyze data in new ways. I believe that we’ll find that broadband discrimination against low-income neighborhoods runs a lot deeper than the way we’ve been thinking about it. My guess is that if we map all of the infrastructure related to broadband we’d see firm evidence of the infrastructure differences between poor and more affluent neighborhoods.

I am sure that if we could gather the facts related to the age of the wires, poles, and other infrastructure, we’d find the infrastructure in low-income neighborhoods is significantly older than in other neighborhoods. Upgrades to broadband networks are usually not done in a rip-and-replace fashion but are done by dozens of small repairs and upgrades over time. I also suspect that if you could plot all of the small upgrades done over time to improve networks, you’d find more of these small upgrades, such as replacing cable company power taps and amplifiers, to have been done in more affluent neighborhoods.

We tend to think of broadband infrastructure as the network of wires that brings fast Internet to homes, but modern broadband has grown to be much more than that, and there is a lot of broadband infrastructure that is not aimed at home broadband. Broadband infrastructure has also come to mean small cell sites, smart grid infrastructure, and smart city infrastructure. I believe that if we could map everything related to these broadband investments we’d see more examples of discrimination.

Consider small cell sites. Cellular companies have been building fiber to install small cell sites to beef up cellular networks. I’ve never seen maps of small cell installations, but I would wager that if we mapped all of the new fiber and small cell sites we’d find a bias against low-income neighborhoods.

I hope one day to see an AI-generated map that overlays all of these various technologies against household incomes. My gut tells me that we’d find that low-income neighborhoods will come up short across the board. Low-income neighborhoods will have older wires and older poles. Low-income neighborhoods will have fewer small cell sites. Low-income neighborhoods won’t be the first to get upgraded smart grid technologies. Low-income neighborhoods won’t get the same share of smart city technologies, possibly due to the lack of other infrastructure.

This is the subtle discrimination that the FCC isn’t going to find in their docket because nobody has the proof. I could be wrong, and perhaps I’m just presupposing that low-income neighborhoods get less of every new technology. I hope some smart data guys can find the data to map these various technologies because my gut tells me that I’m right.

Price discrimination has been around for a long time, but I think there is evidence that it’s intensified in recent years. I first noticed price discrimination in the early price wars between the big cable companies and Verizon FiOS. This was the first widespread example of ISPs going head-to-head with decent broadband products where the big differentiator was the price.

I think the first time I heard the term ‘win-back program’ was related to cable companies working hard not to lose customers to Verizon. There are stories in the early days of heavy competition of Comcast keeping customers on the phone for a long time when a customer tried to disconnect service. The cable company would throw all sorts of price incentives to stop customers from leaving to go to Verizon. Over time, the win-back programs grew to be less aggressive, but they are still with us today in markets where cable companies face stiff competition.

I think price competition has gotten a lot more subtle, as witnessed by a recent study in Los Angeles that showed that Charter offers drastically different online prices for different neighborhoods. I’ve been expecting to see this kind of pricing for several years. This is a natural consequence of all of the work that ISPs have done to build profiles of people and neighborhoods. Consumers have always been leery about data gathered about them, and the Charter marketing practices by neighborhood are the natural endgame of having granular data about the residents of LA.

From a purely commercial viewpoint, what Charter is doing makes sense. Companies of all sorts use pricing to reward good existing customers and to lure new customers. Software companies give us a lower price for paying for a year upfront rather than paying monthly. Fast food restaurants, grocery stores, and a wide range of businesses give us rewards for being regular customers.

It’s going to take a whistleblower to disclose what Charter is really doing. But the chances are it has a sophisticated software system that gives a rating for individual customers and neighborhoods based on the likelihood of customers buying broadband or churning to go to somebody else. This software is designed to offer a deeper discount in neighborhoods where price has proven to be an effective technique to keep customers – without offering lower prices everywhere.

I would imagine the smart numbers guy who devised this software had no idea that it would result in blatant discrimination – it’s software that lets Charter maximize revenue by fine-tuning the price according to a computer prediction of what a given customer or neighborhood is willing to pay. There has been a lot of speculation about how ISPs and others would integrate the mounds of our personal data into their businesses, and it looks like it has resulted in finely-tuned price discrimination by city block.

Is There a Fix for Digital Discrimination?

The big news in the broadband industry is that we are in the process of throwing billions of dollars to solve the ultimate case of economic discrimination – the gap between urban and rural broadband infrastructure. The big telcos completely walked away from rural areas as soon as they were deregulated and could do so. The big cable companies never made investments in rural areas due to the higher costs. The difference between urban and rural broadband networks is so stark that we’ve decided to cure decades of economic discrimination by throwing billions of dollars to close the gap.

But nobody has been seriously looking at the more subtle manifestation of the same issue in cities. The FCC is only looking at digital discrimination because it was required by the Infrastructure Act. Does anybody expect that anything will come out of the stories of discrimination? ISPs are going to say that they don’t discriminate. If pinned down, they will say that what looks like discrimination is only the consequence of them making defensible economic decisions and that there was no intention to discriminate.

Most of the discrimination we see in broadband is due to the lack of regulation of ISPs. They are free to chase earnings as their top priority. ISPs have no regulatory mandate to treat everybody the same. The regulators in the country chose to deregulate broadband, and the digital discrimination we see in the market is the direct consequence of that choice. When AT&T was a giant regulated monopoly we required it to charge everybody the same prices and take profits from affluent customers to support infrastructure and prices in low-income neighborhoods and rural places. Regulation wasn’t perfect, but we didn’t have the current gigantic infrastructure and price gaps.

If people decide to respond to this FCC docket, we’ll see more evidence of discrimination based on income. We might even get some smoking gun evidence that some of the discrimination comes from corporate bias based on race and other factors. But discrimination based on income levels is so baked into the ways that corporations act that I can’t imagine that anybody thinks this docket is going to uncover anything we don’t already know.

I can’t imagine that this investigation is going to change anything. The FCC is not going to make big ISPs spend billions to clean up broadband networks in low-income neighborhoods. While Congress is throwing billions at trying to close the rural broadband gap, I think we all understand that anywhere that the big corporations take the rural grant funding that the infrastructure is not going to be maintained properly and that in twenty years we’ll be having this same conversation all over again. We know what is needed to fix this – which is regulation that forces ISPs to do the right thing. But I doubt we’ll ever have the political or regulatory will to force the big ISPs to act responsibly.

Categories
Regulation - What is it Good For? The Industry

Counting Broadband Locations

All of the discussion of the FCC maps lately made me start thinking about broadband connections. I realized that many of my clients are providing a lot of broadband connections that are not being considered by the FCC maps. That led me to think that the old definition of a broadband passing is quickly growing obsolete and that the FCC mapping effort is missing the way that America really uses broadband today.

Let me provide some real-life examples of broadband connections provided by my clients that are not being considered in the FCC mapping:

  • Broadband connections to farm irrigation systems.
  • Broadband to oil wells and mining locations.
  • Broadband to wind turbines and solar farms.
  • Fiber connections to small cell sites.
  • Broadband electric substations. I have several electric company clients that are in the process of extending broadband to a huge number of additional field assets like smart transformers and reclosers.
  • Broadband to water pumps and other assets that control water and sewer systems.
  • Broadband to grain elevators, corn dryers, and other locations associated with processing or storing crops.
  • I’m working with several clients who are extending broadband for smart-city applications like smart streetlights, smart parking, and smart traffic lights.
  • Broadband to smart billboards and smart road signs.
  • Broadband for train yards and train switching hubs.
  • There are many other examples, and this was just a quick list that came to mind.

The various locations described above have one thing in common. Most are locations that don’t have a 911 street address. As such, these locations are not being considered when trying to determine the national need for broadband.

A lot of these locations are rural in nature – places like grain elevators, mines, oil wells, irrigation systems, wind turbines, and others. In rural areas, these locations are a key part of the economy, and in many places are unserved or underserved.

We are putting a huge amount of national energy into counting the number of homes and businesses that have or don’t have broadband. In doing so, we have deliberately limited the definition of a business to a place with a brick-and-mortar building and a 911 address. But the locations above are often some of the most important parts of the local economy.

I’ve read predictions that say in a few decades there will be far more broadband connections to devices than to people, and that rings true to me. I look around at the multiple devices in my home that use WiFi, and it’s not hard to envision that over time we will connect more and more locations and devices to broadband.

After a decade of talking about the inadequate FCC broadband maps, we finally decided to throw money at the issue and devise new maps. But in the decade it took to move forward, we’ve developed multiple non-traditional uses for broadband, a trend that is likely to expand. If we are really trying to define our national need for broadband, we need to somehow make sure that the locations that drive the economy are connected to broadband. And the only way to do that is to count these locations and put them on the broadband map, so somebody tries to serve them. The current maps are doing a disservice by ignoring the huge number of these non-traditional broadband connections.

Categories
The Industry

The Downside to Smart Cities

I read almost daily about another smart city initiative somewhere in the country as cities implement ideas that they think will improve the quality of life for citizens. I just saw a statistic that says that over two-thirds of cities have now implemented some form of smart city technology. Some of the applications make immediately noticeable differences like smart electric grids to save power, smart traffic lights to improve traffic flow, and smart streetlights to save electricity.

But there are a few downsides to smart city technology that can’t be ignored. The two big looming concerns are privacy and security. There was an article in Forbes earlier this year that asked the question, “Are Privacy Concerns Halting Smart Cities Indefinitely?” Citizens are pushing back against smart city initiatives that indiscriminately gather data about people. People don’t trust the government to not misuse personal data.

Some smart city initiatives don’t gather data. For instance, having streetlights that turn off when there is nobody in the area doesn’t require gathering any data on people. But many smart city applications gather mountains of data. Consider smart traffic systems which might gather massive amounts of data if implemented poorly. Smart traffic systems make decisions about when to change lights based upon looking at images of the cars waiting at intersections. If the city captures and stores those images, it accumulates a massive database of where drivers were at specific times. If those images are instantly discarded, never stored and never available for city officials to view then a smart traffic system would not be invading citizen privacy. But the natural inclination is to save this information. For instance, analysts might want to go back after a traffic accident to see what happened. And once the records are saved, law enforcement might want to use the data to track criminal behavior. It’s tempting for a city to collect and store data – all for supposedly good reasons – but eventually, the existence of the data can lead to abuse.

Many people are very leery of systems that capture public video images. If you look at smart city sales literature, it’s hard to find sensor systems that don’t toss in video cameras as part of any street sensor device. I just saw a headline saying that over 400 police departments now partner with Ring, the video cameras people install at their front door – which allow police to have massive numbers of security cameras in a city. It’s incredibly easy for such systems to be abused. Nobody is uncomfortable with using surveillance systems to see who broke into somebody’s home, but it’s highly disturbing if a policeman is using the same system to stalk an ex-wife. Video surveillance isn’t the only sensitive issue and smart city technology can gather all sorts of data about citizens.

What I find scarier is security since smart city systems can be hacked. Security experts recently told Wired that smart city networks are extremely vulnerable to hacking. Municipal computer systems tend to be older and not updated as regularly. Municipal computer systems have the same problems seen in corporations – weak passwords, outdated and ignored security patches, and employees that click on spam emails.

Smart city networks are more vulnerable to attack than corporate networks that sit behind layered firewalls because a smart city network can be attacked at the sensor edge devices. It’s well known that IoT devices are not as rigorously updated for security as other components of computer networks. I’ve seen numerous articles of hackers who were able to quickly defeat the security of IoT devices.

While there might be a concern that city employees will abuse citizen data there is no doubt that hackers will. It’s not hard to envision hackers causing mischief by messing with traffic lights. It’s not hard to envision terrorists paralyzing a city by shutting down everything computer-related.

But the more insidious threat is hackers who quietly gain access to city systems and don’t overtly cause damages. I have one city client that recently found a system they believe has been compromised for over a decade. It’s not hard to envision bad actors accessing video data as a tool to use for burglary or car theft. It’s not hard to imagine a bad actor selling the data gathered on city networks to players on the dark web.

I’m not against smart city technology, and that’s not the point of this blog. But before a city deploys networks of hundreds of thousands of sensors, they need to have planned well to protect citizen data from misuse by city employees and by abuse from hackers. That sounds like a huge challenge to me and I have to wonder how many cities are capable of doing it right. We’ve seen numerous large corporations get hacked. Smart city networks with huge numbers of sensors are far less secure and look to be an open invitation to hackers.

Categories
The Industry

The Pushback Against Smart Cities

If you follow the smart city movement in the US you’ll quickly see that Kansas City, Missouri touts itself as the nation’s smartest city. The smart city movement got an early launch there when the City was announced as the first major market for Google Fiber. That gigabit fiber network attracted numerous small tech start-ups and the City also embraced the idea of being a technology leader.

The city’s primary smart city venture so far has been to bring smart city technology to a 54-block area in downtown. But this area only covers about 1% of the total area of the City. The City is currently contemplating expanding the smart city into the neglected east side neighborhoods near downtown. This is an area with boarded up storefronts and vacant lots, and the hope is that investing in smart city will bring a boost to this area as a way to kick-start economic development.

So far the primary smart city applications include smart parking, smart intersections, smart water meters and smart streetlights. The city also installed video surveillance cameras along the 2.2-mile downtown corridor.  The existing deployment also includes public WiFi provided through 25 kiosks placed throughout the smart city neighborhood. As of last fall there had been a reported 2.7 million log-ins to the WiFi network.

In the east side expansion WiFi will take on a more significant role since it’s estimated that only 40% of the residents in that area have home broadband today – far below the national average of 85%. The city is also looking to implement a rapid transit bus line into the east side as part of the smart grid expansion.

The new expansion into the east side is slated to have more surveillance including new features like gun shot detectors. There has been public fear voiced that this system can be used to disadvantage the largely minority population of the area.

The biggest hurdle to an expanded smart city services is money. The initial deployment was done through a public-private partnership. The city contributed $3.7 million, which it largely borrowed. Sprint, which manages the WiFi network contributed about $7 million and Cisco invested $5 million. The cost to expand the smart city everywhere has been estimated to cost half a billion.

It is the public-private partnerships that bring a troublesome aspect to the smart city concept. It’s been reported that Sprint collects data from those who log in to the free WiFi network – information like home zip code and results of Internet searches. It’s also been reported that Sprint can track people who have once subscribed to the service, even if they don’t log in. Sprint won’t say how it collects and uses customer data – but as we are learning throughout the tech world, it is the monetization of customer data that fuels many ISPs and online services.

There is also growing public concern about surveillance cameras. It’s starting to become clear that Americans don’t want to be tracked by cameras, especially now with the advent of decent facial recognition technology. We saw Seattle have to tear down a similar surveillance network before it ever went into service. We’re seeing huge pushback in Toronto about a proposed smart city network that includes surveillance.

We only have to look at China to see an extreme example of the misuse of this technology. The country is installing surveillance in public places and in retail areas and tracks where people are and what they do. China has carried this to such an extreme that they are in the process of implementing a system that calculates a ‘citizen score’ for every person. The country goes so far as to notify employers of even minor infractions of employees like jaywalking.

It’s going to be an uphill battle, perhaps one that never can be won for US cities to implement facial recognition tracking. People don’t want the government to be tracking where they are and what they do every time they go out into public. The problem is magnified many times when private companies become part of the equation. As much as the people in Kansas City might not fully trust the City, they have far less reason to trust an ISP like Sprint. Yet the smart city networks are so expensive it’s hard to see them being built without private money – and those private partners want a chance to get a return on their investment.

Categories
The Industry What Customers Want

Smart Cities and Fiber

I’ve noticed that a lot more cities are talking about becoming ‘smart cities.’ Only a few years ago this was something that only NFL cities talked about, but now I see it as a goal for much smaller cities. ‘Smart city’ is an interesting concept. If you listen to the various vendors pushing the idea this means investing in massive amounts of sensors and the computing power to make sense of them. But there are also a lot of lower-tech ideas that fit under this same umbrella.

I’ve had discussion with folks at cities who think that they need fiber in order to have a smart city. Nobody is a bigger proponent of fiber than I am, but fiber is not necessarily needed for many of the concepts that are part of this high-tech vision.

Having smarter traffic flow is generally at the top of everybody’s list. It’s common sense that having vehicles needlessly waiting for lights wastes fuel and wastes time. Smarter traffic lights in cities would improve the quality of life and the economy. A decade ago a lot of cities built fiber networks just to provide a real-time connection to each traffic signal. Those fiber networks allowed the city to change signal timing in reaction to emergencies and similar events, but the whole effort is largely still manual.

But with AI starting to become a realistic technology it looks like truly smart traffic lights are a possibility in the near future. A smart traffic system could change lights on the fly in response to real-life traffic to reduce the average time that vehicles wait for a green light. But the question that must be asked is if this really requires fiber? A decade ago it did. Fiber was needed just to provide the traffic cameras needed to allow somebody at traffic headquarters to eyeball the situation at a given intersection.

But we are now seeing a revolution in sensing devices. We are not too many years removed from the big push to do all heavy-computing in the cloud. A decade ago the vision was that a smart traffic light system would rely on cloud computing power. But faster computers have now reversed that trend and today it makes more sense to put smart computers at the edge of network. In the case of traffic lights, smart computers at the edge reduces the need for bandwidth. Sensors at an intersection no longer need to broadcast non-stop and only need to relay information back to the central core when there is some reason to do so.

For example, one of the uses of a smart traffic system is to identify problem intersections. Sensors can be programmed to record every instance when somebody runs a red light or even a late yellow light and this can alert authorities to problems long before a tragic accident. But these sensors only need to send data when there is an actionable event, and even that doesn’t require a gigantic burst of data.

The same goes for smart traffic control. The brains in the device at an intersection can decide to allow for a longer green for a turn lane if there are more cars than normal waiting to turn. That doesn’t need a big continuous bandwidth connection. The city will want to gather data from intersections to know what the devices are doing, but with smart edge devices a wireless connection provides adequate broadband and a lower cost solution for data gathering.

This same trend is happening with other kinds of sensors. Sensors that listen for gunshots, smart grid sensors used to monitor water and electric networks, and smart sensors used to provide smarter lighting all can be done wirelessly and do not need a fiber connection.

The real purpose behind the concept of a smart city is to provide better government service to constituents. Many of the best ideas out there don’t involve much bandwidth at all. For example, I recently watched a demo of a system in a mid-western city that allows citizens to see, in real time, the location on a map all of the snow plows and trash trucks operating in the city – much like is done when you can see a Lyft ride coming to pick you up. This will drastically cut down on calls during snowstorms since citizens can see a plow making its way towards their street. (And watching the plow inch towards you on a snowy day is good entertainment!)

Cities are undertaking all sorts of other initiatives to improve quality of life. I see cities working on computer systems that put all government forms and processes online, making it easier to get a permit or to report a problem to the city. Cities are reducing pollution by passing ordinances that promote roof-top gardens, that require that new high-rises that are energy self-sufficient and that promote safe bicycling.

There are still big corporations out pitching the expensive smart city vision. But there are now smaller boutique smart city vendors that working towards more affordable and reasonably-priced sensors to spread around a city.

Like anyone who lives in a city I would love to see my city implement smart city ideas that improve the quality of life. But as much as I am a fiber-proponent, I am finding it hard to make a case that a lot of urban fiber is needed to implement the best smart-city ideas.

Exit mobile version