Heute 29

Gestern 88

Insgesamt 39186561

Freitag, 10.07.2020
eGovernment Forschung seit 2001 | eGovernment Research since 2001

The smart cities movement has evolved from sensors on light poles aimed at improving traffic flow to connected platforms that are practical, efficient and equitable for government, businesses and residents.

Like so many American cities these days, Pittsburgh finds itself suffering from significant growth in traffic and road congestion. By 2017, drivers were spending an extra 81 hours commuting to work each year. To ease the problem, the city worked with Carnegie Mellon University to build a traffic signal system that ran on artificial intelligence instead of relying on pre-programmed signal cycles.

The results were soon apparent. For the initial 50-intersection project, the system reduced travel time by 25 percent, braking by 30 percent and idling by more than 40 percent. The AI software detects traffic and creates a predictive model that generates a signal timing plan in real time.

While drivers were happy, pedestrians let the project team know that they felt left out of the picture. So, the researchers responded by tweaking the system to minimize wait time for pedestrians at lights. Meanwhile, researchers and students at Carnegie Mellon set to work on a side project to make a mobile phone app to communicate with the lights for people with disabilities who need more time to cross the street.

The project in Pittsburgh is more than just a smart city initiative. It’s an example of how the movement has evolved. It tackles an ordinary problem using the latest technology in an innovative way. There are multiple players involved, including a startup — Rapid Flow Technologies, which was spun out of Carnegie Mellon — and it was built to scale up. Eventually, more than 600 intersections throughout Pittsburgh could be using the technology. Finally, the solution aims to be equitable, affecting the entire community, including those who don’t have cars but use the streets anyway.

“Smart city projects are becoming less fantastical, less sci-fi type and much more practical,” said Bob Bennett, founder and principal of B Squared Civic Solutions and former chief innovation officer of Kansas City, Mo.

Start with People

The rapid urbanization of cities in America and around the globe coupled with the rise of technology in the past 20 years has presented technologists and policymakers with a unique opportunity to try to fix the problems with life in cities — traffic, pollution, crime — using hardware and software. Cities like Santander, Spain, and Rio de Janeiro became showcases for sensors, cameras and other devices linked to networks, servers and dashboards that would collect, measure and analyze reams of data. The goal was to find new ways to figure out a city’s problems and deliver quick answers.

The problem was that technology began to dominate the smart city conversation. The idea was that the municipal leaders needed tools in the form of the Internet of Things to resolve urban issues, because the problems were too complex for humans to track and make decisions about in a timely fashion. Despite a wave of marketing hype and numerous pilot projects involving just about every kind of technology available, few cities were becoming truly smart and connected.

The McKinsey Global Institute released a report in 2018, Smart Cities: Digital Solutions for a More Livable Future, that explained the existing problem with smart cities: “After a decade of trial and error, municipal leaders are realizing that smart city strategies start with people, not technology. ‘Smartness’ is not just installing digital interfaces in traditional infrastructure or streamlining city operations. It is about using technology and data purposefully to make better decisions and deliver a better quality of life.”

That’s not to say technology is taking a back seat to making cities smarter. Far from it. But the human equation has grown in importance. “Equity is an underlying requirement when it comes to saying which smart city use case will have the broadest impact during the evaluation process,” said Michele Pelino, a principal analyst with Forrester Research. “Cities want projects that give them the most value.”

On a more fundamental level, cities are accepting the fact that smart city projects often start within a silo and that will remain the case. “The siloing of projects remains consistent because that’s where the funding is,” said Bennett.

What has changed is that while a project with sensors on a lightpole might start within one city agency, the solution is embraced interdepartmentally. “There is a better appreciation of what the impact is going to be across the enterprise, instead of being a project where the benefits are also siloed,” Bennett explained.

In other words, the business model for smart city projects is improving. Instead of sticking water sensors in the ground of city parks because technology can measure whether or not the sprinklers should be turned on, they are focusing on using sensors that improve how the city’s water treatment system operates, benefiting the entire community.

Cities have also begun to recognize that their infrastructure can become platforms. Streetlights, for example, are no longer performing one smart task, but several. While the more advanced cities note this opportunity up front, most municipalities realize the importance of building a platform later on, according to Pelino. “Oftentimes, it happens after the fact, when the city has a number of these individual elements in place and there is a complexity that needs to be simplified,” she said. “The insight captured becomes valuable when it can be shared in city agencies, which allows for the sharing of information and intelligence.”

Smart Starts at the Edge

In 2019, the Center for Digital Government* conducted its annual Digital Cities survey and found several smart city tech trends underway. Cities had boosted the amount of technology they were using in key areas since the 2018 survey and expected to continue that expansion.

The first generation of smart city projects focused on figuring out how to put sensors in the community, network them together and feed the data they collected back into data centers where the information could be analyzed and intelligence extracted. But now cities have recognized there’s more value in having the data processed and analyzed in the field. Edge computing has driven computing power out into the streets.

“To understand the value happening in an individual element, whether it’s a bus traveling down a street or a sensor on a street pole, the processing of the information has to happen at the location of the asset because that’s where you need to make the decision,” Pelino said.

Chicago has long been a testbed for practical uses of smart city technology. For years, the Array of Things (AoT), a joint research project between scientists, universities and government, has looked into ways to improve how IoT and artificial intelligence can help cities connect and solve problems, using the streets of the Windy City as its testing ground.

In the early days, AoT focused on how to make sensors work better in the field, in all kinds of weather conditions, as well as determine the best applications for a city’s needs. Today, the focus is squarely on edge computing, according to Charlie Catlett, one of the founding researchers of AoT. “Our research is about what is possible using technology, while holding down costs and protecting the public’s privacy,” he said.

The way to do that is to build software-based sensors that can conduct a variety of tasks, using cameras, microphones and other devices that can collect vast amounts of data and analyze them at the location. AoT is now working on a new project called SAGE, which will explore techniques for applying machine learning algorithms to data from intelligent sensors and run reusable software programs within the embedded computer, and transmit the results over the network to central computer servers.

In other words, the sensor is the computer. “We’re at the tipping point with edge computing,” Catlett said. Imagine sensors that can decide on their own when they should turn on their cameras to analyze traffic, test the air for particulate matter or look for evidence of water pooling on the streets after a flash rainstorm. “We now have sensors that are smart enough to know what they should do and when they should do it, and then make decisions based on the information gathered at the location,” he said. “We are going toward autonomous measurement.”

Part of the problem with early smart city projects was that many were built by companies that were proprietary in design, limiting their flexibility and capability, according to Catlett, who is currently on sabbatical from AoT and is a senior research scientist at Discovery Partners Institute, a tech talent development organization in partnership with universities. “The hardware for these experiments was out ahead of the software,” he said. For smart city technology to succeed, you need to flip that equation.

That’s the thinking behind the SAGE project: intelligent sensor nodes that support machine learning algorthims that have been created using open source code, providing an open architecture, so that cities can build an array of projects on one platform.

It’s About Performance, Efficiency and Equity

So what are cities looking for in 2020 when it comes to being smart and connected? Early applications included self-driving vehicles, drones, cameras (on bodies and in streetlights) and sensors to measure every conceivable function. But today, it’s about being pragmatic with smart technology. “Smart city applications are staying within their foundational elements, which have always been transportation-oriented systems, utilities, including water and power,” says Bennett.

Public safety is another active area, along with bringing connectivity to communities — think free Wi-Fi in parks and downtowns. “I’d also include waste management; this one comes up quite a lot today,” Pelino said. “Cities want to better manage their waste; they want connected trash cans that tell workers which bins need to be unloaded.”

Along with a return to the basics, smart city projects are also becoming more focused on performance, efficiency and, perhaps most important, equity. The city of San Diego has installed roughly 3,000 smart streetlamps over the past few years, making it one of the largest smart platforms for an urban area in North America. The promise was that by tracking the movement of cars and people, the city would be able to ease its notorious traffic and parking problems.

However, what was envisioned by city officials as a highly beneficial use of technology was seen as something more sinister by the residents, especially after they found out the police would have access to the camera footage. When San Diego’s newspaper analyzed data from the project, it found a slightly higher concentration of smart streetlights in white neighborhoods, leading to charges of tech elitism, while others read the data to indicate that too many cameras were near neighborhoods of color, leading to charges of racial bias in how decisions were made about sensor location.

The lesson learned in San Diego is that deploying smart city technology is not as simple as it seems. But cities are learning how to become more equitable. Projects still start in a stovepipe, said Bennett, but the evaluation and use by local departments is based on the broader equity it brings.

Smart city projects are also evolving into solutions that deliver outcomes. It’s not enough to have technology that can tell a sanitation crew a certain bin is overflowing and needs to be emptied. It has to show outcomes, such as fewer complaints to the city’s 311 system about trash in the streets, Pelino said. “If you can get fewer citizens calling the city to empty an overflowing trash bin, that can be a metric that can show the value of the application,” she explained. “This leads to more cost-efficiency measures on how workers are spending their time; it can even lead to fewer calls about rats in alleys, because cleaner streets can lead to reductions in the rat population.”

The Stakeholder, Worker Equation

Collecting data from thousands of sensors, cameras and microphones, analyzing the input, and then figuring out the best ways to fix what ails a city isn’t easy or simple, no matter how sophisticated the technology. Nor is it cheap. It’s not surprising, then, that while cities have embraced the concept of being smart and connected, they have stumbled in execution.

Part of the problem has been having the right stakeholders in place to say what the city needs and why. Sometimes the choice is forced by a regulatory requirement. But more often, it takes the right kind of leader to ask CIOs to think differently in how to apply technology and to tell workers they need to do their jobs differently because of sensors in the roads or their vehicles.

Indeed, city workers are key if smart cities are to succeed. They have to develop different skill sets, accept that data may require them to do their job differently, or fix something based on an algorithm rather than driving by and seeing for themselves whether something needs fixing.

Technology too can be a roadblock. Adding a new system to an existing platform without interrupting a service that citizens have come to rely on can be disruptive, warned Bennett. Ultimately, if smart cities are to evolve and grow, certain changes in behavior, management and organization will be needed. “It’s about understanding the value of the data in the dashboard,” said Forrester’s Pelino, “and prioritizing what needs to be done.”

---

Autor(en)/Author(s): Tod Newcombe

Quelle/Source: Government Technology, June 2020

Bitte besuchen Sie/Please visit:

Zum Seitenanfang