SAFETY AND SECURITY

The great migration is on. By 2050, two out of every three people will live in cities or urban centers. Keeping an eye on the future, cities are working hard to stretch their budgets to make significant and necessary investments in infrastructure.

Unfortunately, they frequently lack the know-how to integrate it all and make the most of their investment. “Some of the infrastructure is a year old, some are 30 years old, so integrating existing systems to get a true understanding of what’s happening is a challenge,” says Matt Powell, Managing Director, North America for Intelligent Security Systems (ISS), a supplier of intelligent video management systems and analytics. In addition, Powell adds, “there’s never enough people or money to solve every challenge, so cities have to figure out how to leverage technology to empower city workers and residents.”

Rounding up disparate infrastructure systems to generate usable data for AI in city planning is another problem. Addressing this challenge—the paucity of data—is where AI-driven video analytics from ISS can help. The company provides notifications about events unfolding in real time, and identifies long-term trends about pedestrian behavior, vehicle crashes, traffic jams, and more.

“Better environmental data, combined with video analytics, provides a lot of information that can help in terms of cities’ safety and security planning.” Powell says.

“Better environmental #data, combined with #VideoAnalytics, provides a lot of information that can help in terms of cities’ #safety and #security planning.” — Matt Powell, Intelligent Security Systems via @insightdottech

Walking the Walk: Pedestrian Safety

City safety and security can mean a lot of different things, but decreasing the number of pedestrian incidents—2021 registered 7,000 pedestrian deaths in the United States—is an important concern. ISS is part of an ecosystem of Intel partners working together to help cities realize their goals for VisionZero, a strategy to eliminate all traffic fatalities and severe injuries, while increasing safe, healthy, equitable mobility for all.

Focusing on crosswalks is one component of this equation. A pleasant bonus: Getting pedestrianism right is a path for local governments to achieve ambitious carbon reduction goals.

Lighting on crosswalks frequently falls short of keeping pedestrians safe, Powell says. Earnest efforts, adopting a range of approaches, are underway. Singapore, for example, is bathing sidewalk curbs in red light so pedestrians on their phones can notice danger more easily. Flashing lights on crosswalks are becoming more ubiquitous in the United States. But these safety measures continue to put the burden on pedestrians for their own safety.

Drivers, too, can miss pedestrians as they make corner turns or are in poorly lit conditions. At the same time, too much illumination in crosswalks can be blinding.

ISS’s SecurOS Soffit Intelligent system addresses these problems by using dynamic illumination that tracks pedestrians in sidewalks. The spotlight consists of a group of LED modules that illuminate different sections of the pedestrian crossing. The system, which also includes an IP camera and video analytics controller, delivers low-powered static lighting of the sidewalk in standby mode.

AI-driven algorithms detect pedestrians from any angle of approach and activate lights to shine on the specific portion of the crosswalk the pedestrian occupies. As the pedestrian moves, the corresponding LED module switches on, too. The system can provide dynamic lighting for multiple pedestrians on a sidewalk at a time, approaching from different directions and at different speeds.

Analytics-First Approach to Safety and Security

ISS uses edge AI algorithms to process video. “The analytics require powerful computing, which is where we utilize Intel as a partner,” Powell says. “Intel has worked with us to develop processing kits that allow us to go out and install these systems – mobile with a small footprint – and put it in GPU cores.” ISS extensively trains its AI algorithms on neural network models to detect and classify road users, whether they’re vehicles, pedestrians, or animals.

Most cities do not need an extensive infrastructure reboot to incorporate the Soffit solution. By this point they have the IP cameras with servers feeding back to a command center of some type. And ISS works along with integration partners to train them on solutions. “We don’t just drop it off on the doorstep,” Powell says.

Powell stresses that ISS is “an analytics company first and a video management platform and decision management system second,” which means it’s extremely receptive to integrations with other components. An open API and easy-to-use software development kit help. “The future is non-proprietary. We do have our own system, but we have always recognized that we’re a value-add to any module that’s deployed,” Powell says.

The analytics-first approach enables ISS to do more with video. Intelligence derived from Soffit, for example, can forecast peak usage times, which direction people come from, and when. The solution also documents near misses. “When such information is fed to a planning department, they can understand what changes need to be made to the environment – does the area need a stop sign, is it a blind curve…?” Powell says.

Shaping IoT and Smart Cities

Better planning will be key in shaping the smart city of the future. And video will shape these new landscapes, Powell says. “To accommodate smart vehicles and smart communications, we’ll need video analytics to see how these systems play out. They will have an important role in understanding the changes we need to make to create a safer, smarter environment, one that has better traffic flow and fewer casualties.”

This article was edited by Georganne Benesch, Associate Editorial Director for insight.tech.