New deployment combines edge AI, radar, computer vision and connected infrastructure to explore how intersections can become intelligent perception nodes for cities and future mobility
What happens when a traffic intersection becomes more than traffic signals, cameras and roadside equipment?
At Curiosity Lab, we’re exploring that exact question by partnering with LEOTEK to deploy its Interlux™ AI Traffic Signal System at a live intersection in Peachtree Corners.
The deployment brings together radar, computer vision, edge AI computing and connected-vehicle technologies to create a more complete, real-time understanding of what is happening across an intersection.
For Curiosity Lab, this goes far beyond testing another traffic tool. It highlights a larger shift we are seeing across smart city development: physical infrastructure is becoming an intelligence layer for the city.
Giving the Intersection Its Own Intelligence
Today’s vehicles are equipped with sophisticated cameras, radar and sensors. But even the most advanced vehicle has limits.
Buildings, trucks, other vehicles and the geometry of an intersection can create blind spots. A pedestrian can be hidden behind a vehicle, or a turning vehicle may not have visibility around an obstruction. In these moments, intelligence located outside the vehicle becomes critical.
Interlux is designed to bring that missing perspective directly to the intersection.
By combining camera and radar data with edge AI processing, the system detects and interprets movement from vehicles, pedestrians, and cyclists in real time. Rather than routing data to a distant cloud server, processing happens instantly at the roadway edge.
The longer-term potential is significant: an intersection can potentially become an additional set of “eyes” for connected and autonomous vehicles, while simultaneously providing cities with a richer data traffic flow, safety and physical operations.
“What makes this interesting to us is that the intersection is no longer just managing traffic signals. It can become part of the intelligence infrastructure of the city,” said Pouyan Azizpour, Innovation Program Manager at Curiosity Lab. “A vehicle has its own perception, but infrastructure can see things the vehicle cannot. Testing how those two worlds eventually work together is exactly the type of work Curiosity Lab was built to support.”
Testing Physical AI Where It Matters: In the Physical World
AI is rapidly moving beyond software applications and stepping out onto our roads, robots, vehicles and physical systems.
Interlux is a prime example of that shift.
LEOTEK’s system relies on NVIDIA Jetson Orin edge computing, radar-camera sensor fusion and connected transportation tech to process activity at the intersection. LEOTEK describes the architecture as a way to turn traditional roadside equipment into dynamic sensing and computing nodes.
The deployment generates a live digital representation of the intersection, providing another layer of information about how road users move and interact.
That opens up exciting possibilities, including intersection safety analysis, traffic-flow optimization, pedestrian and cyclist awareness, emergency-vehicle priority routing and enhanced vehicle to infrastructure communication.
By deploying it at Curiosity Lab, we can see how these capabilities perform outside of a controlled lab environment and inside a busy, functioning city intersection.
A Real Intersection, Not a Controlled Test Track
Curiosity Lab is owned and operated by the City of Peachtree Corners which provides companies with access to real municipal infrastructure, including our three-mile autonomous and connected vehicle test corridor, C-V2X connectivity, intelligent traffic infrastructure, cameras, sensors and other smart city technologies.
Technologies deployed here encounter the same conditions cities deal with every day: shifting traffic volumes, foot traffic, turning vehicles, existing infrastructure, connectivity requirements, power limitations and multi-system coordination.
For LEOTEK, this setup creates an opportunity to validate Interlux under real-world conditions while also exploring how it can interact with the broader technology ecosystem already operating in Peachtree Corners.
For the City, it creates something just as valuable: an opportunity to understand the technology before cities are asked to adopt it at scale.
That is a central part of Curiosity Lab’s mission. We aren’t just a site for product demos; we build an environment where innovators, city departments, infrastructure partners, and tech companies can figure out what real-world deployment actually takes together.
From Traffic Signals to City Intelligence
Intersections may eventually become some of the most valuable computing and sensing locations in a city.
They sit where pedestrians, vehicles, public infrastructure and increasingly connected technologies meet.
Today, most intersection infrastructure performs specific functions independently. The future looks very different: cameras, radar, V2X, edge computing and AI models working together to generate a shared understanding of the physical environment.
That creates capabilities beyond simply changing a traffic light.
Infrastructure will be able to identify emerging safety risks, understand multimodal movement, communicate information to connected vehicles, support digital twins and provide cities with better operational intelligence.
“Deployments like Interlux help us understand what that next generation of infrastructure actually looks like,” Azizpour said. “Instead of talking about the smart intersection as a future concept, we can put the technology on a real intersection and learn from it now.”
The People Behind the Deployment
The deployment has also been a close collaboration between Curiosity Lab, the City of Peachtree Corners, LEOTEK and the City’s infrastructure partners.
David Tzou, AI Solution Architect at LEOTEK, has served as one of the primary technical leads behind Interlux and worked closely with the Curiosity Lab team throughout the deployment process.
LEOTEK’s U.S. Intelligent Transportation Product Manager Eric Zhao and Smart Mobility R&D leadership have also been involved as the teams worked through the product, infrastructure and installation requirements.
The project reflects something we see repeatedly at Curiosity Lab: bringing emerging technology into the real world requires more than innovative hardware or software. It requires technical teams, city operations, public works, infrastructure partners and technology companies working together.
An Award-Winning Platform Comes to Peachtree Corners
Interlux has already received international recognition, including a 2026 iF Design Award for Product Design and a Silver Award in the Smart City Mobility Infrastructure category at the 2025 Edison Awards. LEOTEK is also a member of the NVIDIA Inception program and has continued expanding its work around edge AI and Physical AI for transportation infrastructure.
Yet awards and demonstrations are only part of the journey.
The next question is how technology performs when installed on the infrastructure cities already operate every day.
That’s the part we’re excited to explore.
Building the Next Generation of Connected Infrastructure
LEOTEK joins a growing Curiosity Lab deployment portfolio alongside companies working across intelligent infrastructure, Physical AI, connected mobility, autonomous systems, IoT and other emerging technologies.
Our mission remains simple: give promising technology a path into the real world, give companies meaningful environments to learn and improve, and give cities an earlier understanding of technologies that may shape their future.
With Interlux now at Peachtree Corners intersection, we’re ready to see what happens when intelligence moves beyond the dashboard and directly into the road itself.
It’s a promising step toward safer, more responsive, and better connected cities.