RETAIL

It’s no longer a question of whether we should make our cities “smart,” but it’s now a matter of when and how. Rapid advances in AI, deep learning, and edge computing have the potential to improve our quality of life beyond what was achievable just a decade ago.

And these new smart city technologies aren’t just designed to make citizen life more enjoyable; they’re making it safer, too—and for a fraction of the cost.

Automated Traffic Incident Detection

The Intel^® Distribution of OpenVINO^™ Toolkit helps traffic operators reduce the number of accidents on the road.

For instance, in 2004 after a spurt of traffic accidents, European Union member countries issued minimum road safety requirements for tunnels over 500 meters long. The directive included installation of safety cameras, with the goal of monitoring events like wrong-way drivers, smoke/fire, stopped vehicles, and pedestrians on the roadway.

But the sheer quantity of traffic footage streaming from all those endpoints (potentially hundreds of cameras per tunnel) made manual analysis impossible. In fact, even traditional computer vision camera monitoring couldn’t identify people, objects, and events with enough accuracy to avoid false alarms.

Instead, traffic operators needed AI solutions at the edge to automate traffic incident detection and keep traffic flowing safely.

TRAFFIX.AI, a software solution from mobility analytics company Sprinx uses OpenVINO to run neural networks that do just that. Thanks to superior edge processing, the solution analyzes video feeds and detects everything from wrong-way drivers, slowdowns, and even smoke or fog—in real time.

Enhancing the Spectator Experience

And it’s not just on the roads. Venue organizers use IoT smart sensors and cloud-based technologies to ensure the safety of concertgoers and sporting-event attendees.

With AI platforms like OpenVINO, venue operators can gain accurate, real-time people counts to manage crowds effectively. When facility managers can see where people are gathering at any given time, they can respond quickly to emergency situations. And by analyzing historical data, venues can reconfigure seating, concessioners, and entry points to prevent incidents from happening in the first place.

One solution from IOTech Systems, an edge software provider for the IIoT, leverages the latest Intel technologies around GPU and CPU architectures and the OpenVINO Toolkit to get real-time, actionable crowd data to the people who need it. Organizers can make fast decisions by accessing a simple dashboard from their phone or tablet. And attendees can view open seats on large monitors.

The result? A better experience for spectators, and reduced costs for venues.

AI-Based Video Analytics for Smart City Technology

The use cases for AI-powered video analytics don’t end there. By turning existing CCTV cameras into cost-effective smart sensors, these edge-to-cloud systems can also improve building security or help truckers find a safe place to rest.

AI-based vision systems like the one from Uncanny Vision, an AI video analytics solutions provider, depend on OpenVINO and small embedded processors for lightning-fast processing speeds. And it’s that speed that allows the system to watch and analyze video and inform operators in real time when action is needed. This way, human resources are spared the labor-intensive work of reviewing video streams themselves—and the inevitable errors that follow.

But the best part? False alarms are a rare event. Unlike first-generation sensors that couldn’t tell the difference between a person and an animal, Uncanny’s systems are guaranteed to be 95% accurate. That’s orders of magnitude better than even traditional computer vision systems.

#SmartCities equipped with new #IoT smart sensors and #AI #technology means endless possibilities in creating safe and sustainable cities. @Inteliot via @insightdottech

What’s more, customers and systems integrators don’t need any special knowledge of AI to install the Uncanny system. All they need to do is connect a CCTV camera to one of Uncanny’s Intel processor-based smart boxes and the system is up and running.

AI in Smart Cities Protects Privacy

Importantly, turning security cameras into IoT smart sensors in this way doesn’t breach personal privacy. On the contrary, AI software processes video data at the edge—without recording, storing, streaming, or sharing images over the cloud.

Because the images themselves aren’t needed, all that matters is the software’s ability to identify dangerous security events and escalate them as soon as they happen.

SensingFeeling’s SensorMAX platform does this by using pre-trained models and behavioral analytics to assign a risk index to every camera. Then only potentially risky scenes are displayed to operators—an unaccompanied child in a train station, say, or a crowd gathering in an unusual place. And rather than wasting time scrolling through uneventful camera feeds, the operators are available to respond immediately when they need to.

The SensorMAX platform’s architecture is based on the OpenVINO Toolkit and has a variety of uses beyond smart city applications, including reducing oil and gas accidents.

But improving the spectator experience and keeping people safe are just some of the ways AI is used in smart cities. The potential for innovation is huge, and cities and businesses already capitalize on it.

Smart cities equipped with new IoT smart sensors and AI technology means endless possibilities in creating safe and sustainable cities. See what else AI can do, and start creating your own applications by checking out the Intel Edge AI Certification Program or taking the 30-Day Dev Challenge.

This article was edited by Christina Cardoza, Associate Editorial Director for insight.tech.