This year, city governments globally are expected to spend an estimated $135 billion deploying smart city technology, according to IDC. Yet, despite the billions of dollars invested, American cities still lag behind other global urban centers in using technology to improve the lives of their citizens.

Not surprisingly, the biggest barriers to building smarter U.S. cities have little to do with technology; they revolve around people, processes, and privacy. We spoke with Lin Nease, chief technologist for IoT at Hewlett Packard Enterprise and an HPE fellow, about where America's smart cities are today, the challenges they face, and what the future holds.

First, a basic question. What makes a city 'smart'?

Smart cities are really about the fusion of key services that every city provides and doing it in a more intelligent way to improve the quality of life for citizens.

Where have you seen the greatest advancements in smart city tech?

Probably the biggest opportunity for advancement over the past 15 years has been the installation of LED street lights—and not just because the lights are easier to maintain or more energy efficient. It's because of what else you can do with them. If you're going to the expense of rolling a truck to replace every streetlight, it's reasonable to ask what other technologies you can install alongside them. So we have everything from acoustic gunshot detectors to air quality sensors to security cameras on top of these lamp posts. You've already got a power source and the economic justification for doing it. I'd say that's the biggest breakthrough.

But sensors are only part of the equation, right?

Yes. I think sensors are overrated. The real value comes from leveraging the data that's available from different sources and making intelligent inferences based on that data.

Take COVID, for example. We have process manufacturing customers who must have thousands of employees physically involved in production. They need to make sure employees maintain adequate social distance within a facility. Of course, this can apply to any environment where employees or citizens must be in close proximity, including public works processing plants, critical infrastructure workplaces, and more.

In these environments, using Bluetooth beacons to triangulate each employee's location frequently isn't accurate enough. However, based on historical data, it's possible to use machine learning and statistical analysis to infer when employees are less than six feet apart and then perform automatic contact tracing if someone may have been exposed. If you can do that quickly enough, you can avoid having to shut down or cripple and an entire operation, which could have a major effect on city services. The ability to take imperfect data and use AI to infer conclusions from it will be a huge enabler for smart cities.

Another big breakthrough has been the creation of public-private partnerships, or what I like to call the freshwater-saltwater exchange. Most services provided in smart cities are actually provided by private enterprises. The question then becomes, what kind of data sharing agreements do they have?

What are some of the barriers to rolling out smart services?

They're very similar to the change management challenges enterprises encounter when they embrace digital transformation. Now that you have all this data, what do you do with it? Does it fit into your current processes? If not, you're going to need a process redesign. As we've discovered with a lot of IoT deployments, organizations don't know how processes should change until they've started to implement the technology. It becomes a chicken-and-egg problem.

There are also socioeconomic concerns. Once I've outfitted the sanitation trucks with cameras, I no longer need spotters to make sure drivers don't hit anything when backing up the truck. What happens to them?

How important is edge computing in making cities smarter?

It's critical. Anything related to video processing, like cameras installed in LED lights, needs some form of physically proximate compute power. In its most distributed form, edge compute will be placed in mobile providers' base station towers throughout urban areas. Then the question becomes, how can cities tap into the data that's coming from service providers?

Say I'm a mobile network provider, and I've got half a rack of small servers at 500 locations around the city of San Francisco. I might rent out some of that capacity to software operated by private companies that share data with the city, or to the city itself. And just as multiple carriers share the same towers, these mini-edge computing centers will be rented out to both private interests and public agencies.

Can you give us an example of a smart city project HPE has worked on that stands out to you?

We've been working on a big project in Bhopal, which is a hub for six other cities in Madhya Pradesh, India. We helped create a cloud-based integrated command and control center (ICCC) that pulled data from thousands of sensors, as well as public sources like Google Maps. It allowed the state government to monitor smart lighting systems, traffic, parking, waste management, and water use for more than 20 million residents from one location. Having a single ICCC was much more efficient and saved the local government significant investment. It's part of a larger initiative by the Indian government to create 100 smart cities over the next few years.

It seems like other countries are ahead of the U.S. in many ways.

They are. But one of the reasons smart cities in countries like India and United Arab Emirates have evolved more rapidly has to do with privacy concerns. The fact is, they're less concerned about it than we are. That means they can move faster and introduce more functionality sooner.

In China, the advances are even more profound because they have no scruples about privacy. For example, the Chinese have a social credit system that assigns you points based on your behavior and determines things like your right to travel outside the country. Surveillance cameras using facial recognition in Beijing are extremely efficient and can identify you at walking distance. That leads to some scary Big Brother scenarios.

What's next? How are smart cities likely to evolve?

Using citizens as sensors is one of the most profound areas where smart cities may evolve. How did we catch the Boston bombers? From pictures on people's cell phones. Using citizens as sensors made traffic work better for everyone. That's why Google acquired Waze. I don't go for a run now without first checking PurpleAir.com, which mixes air quality index data from private sensors as well as public.

But that in turn also raises a bunch of privacy concerns. We recently installed a Ring security system that has a view of our entire street. Now, more than 1,300 U.S. police departments can access video feeds directly from Amazon Ring doorbell cameras, which citizens can allow by pushing a button. It's become a bit controversial, especially when you factor in improvements in facial recognition. My camera sees into the street. I don't own the street or the rights to people using the street, but my camera can see them.

So when citizens crowdsource sensor data, who owns it? How do governments get the rights to it? Can they share it with private enterprise? How can they use this data to make better decisions about services that help citizens without violating people's rights? When we can answer those questions is when I think we'll really see smart cities begin to take off in this country.

Dan Tynan