What the PF-Series Means for People Flow Management
System-on-chip (SoC) capabilities from industry leaders such as Synaptics were advancing rapidly. At the same time, global interest in cost-effective physical analytics solutions was growing, as more industries began to recognize the value of real-time analytics and digital twins.
At that moment, Xovis faced two clear paths forward. The first was to develop next-level features for our PC-Series sensors. These sensors were already the industry’s preferred hardware for accurate people counting and tracking, but their platform would require significant effort to support the large-area capabilities we envisaged. The second was to return to the drawing board and design a new hardware solution that leveraged emerging technologies to meet evolving demands for advanced physical analytics.
Had we chosen the first option, I probably wouldn’t be writing this now. Instead, we chose to do what we have always done: develop edge AI-powered tools that make people flow management simpler and more cost-efficient. What follows explains the how and why behind the PF.
Starting from Scratch
The PF-Series is an edge AI people flow management sensor designed to deliver real-time people counting and tracking directly on the device. Edge AI refers to artificial intelligence that runs on the sensor itself, enabling advanced computer vision processing without reliance on cloud infrastructure.
Xovis has been developing purpose-built people-flow management sensors for more than 15 years. We have unique expertise and a leading industry position because of that experience. But people flow management is a fast-moving market, with new players constantly looking to disrupt. Put otherwise: it’s not an industry where you can rest on your laurels.
Our continued success is directly linked to having a strong and talented R&D department, one that is on the cutting edge of computer vision. That team, and the entire Xovis group, has to not only think about what can be done, but also why it should be done.
We usually take a blank slate approach to these questions, working our way from the challenge to the best solution. We did the same here, giving due consideration to a range of technologies—including LiDAR and CCTV—before arriving, again, on 3D stereovision technology.
From that point, and in partnership with Synaptics, we built the PF sensor from the ground up over several iterations across multiple years. By running advanced computer vision models directly on the sensor, the PF-Series delivers real-time people counting and tracking at the edge, reducing latency, bandwidth requirements, and infrastructure complexity.
If we thought that off-the-shelf technology was adequate to solve the challenges that the PF-Series sensors address, we could have saved ourselves a lot of time. But the truth is it’s not, and we are not willing to cut corners just to bring something new to the market. This approach doesn’t always put us at the top of the list of companies that put a premium on novelty, but that’s never been our goal.
What is edge AI in people flow management? Edge AI in people flow management refers to running AI-powered computer vision models directly on intelligent sensors, enabling real-time people counting and tracking without dependence on cloud processing.
Partners on the Edge
“PF-Series is a clear proof point for integrated Edge AI. By unifying AI compute, vision acceleration, and system control on Astra SL1680, we enabled real-time vision at the Edge—and Xovis is proving that architecture at production scale.”
—Vikram Gupta, SVP & GM, Edge Compute & Connectivity Division, Synaptics.
Bringing new, groundbreaking technology to market involves many moving parts. With the PF, Xovis was very fortunate to have forged a strong connection with Synaptics.
Synaptics’ SoC technology delivers a fantastic cost-compute ratio, which is why we selected it for the PF platform, but Xovis also received invaluable guidance from the company throughout the development process. Strong edge AI partnerships such as this accelerate innovation across the entire market. The PF-Series demonstrates how edge AI can transform large-area analytics by delivering production-scale vision processing directly on intelligent sensors.
Why is edge AI important for airports and retail? Edge AI enables faster analytics, lower infrastructure costs, improved data privacy, and scalable deployment across large areas such as airport terminals and retail spaces.
A Broader Horizon
Large-area analytics in the retail and airport management industries is not new, but achieving seamless coverage with nearly 60% fewer sensors is possible because advanced edge AI processing enables broader, more intelligent sensor coverage.
For airport passenger flow management, edge AI enables immediate, on-device analytics that support queue optimization, resource allocation, and terminal operations without sending sensitive data to the cloud. In retail environments, edge AI-powered people counting supports footfall analytics, occupancy monitoring, and store performance optimization in real time.
The PF perfectly fits the intersection of technological capabilities and industry demand that has emerged in recent years. And we at Xovis are excited to help more businesses, more organizations, more retailers, and more airports realize their physical analytics goals in a more cost-efficient manner.
With the PF, we have the perfect platform for developing the next generation of AI-powered, vision-based tools. Some features already in development are so transformative in terms of delivered insights that organizations may initially find it hard to believe what is now possible. The potential is remarkable, and we’re excited to explore these new capabilities with more businesses.
Tags: | People Flow Management | PF-Series | Computer Vision | Real-Time Analytics | Smart Sensors |
TransportationMichael Westermann Discusses APC Trends
- Transportation
Michael Westermann Talks APC Trends
Guides, Airport, Retail, Transportation, Smart buildingsValue of Data Guide