Privacy by Design: How Ouster BlueCity Fuses Native Color Lidar with Edge Privacy

Ouster BlueCity with Rev8 is the industry's first native color lidar detection system, giving traffic engineers up to 500 feet of advance detection and richer situational awareness at intersections. Its edge-processed AI, running on NVIDIA Orin AGX and trained on 4 million labeled objects across 800 sites, delivers 99% stop bar detection accuracy while automatically blurring pedestrians and cyclists to protect identity.

As urban city centers grow and transportation networks become increasingly complex, modern traffic management forces a challenging compromise on smart city planners. Today’s municipalities and departments of transportation (DOTs) require high-fidelity, real-time spatial intelligence to optimize traffic flow and protect vulnerable road users (VRUs). However, capturing this granular level of data frequently triggers significant, valid privacy concerns regarding public surveillance and identity tracking. For today’s transport planners, finding an advanced solution that strikes a balance between robust operational intelligence and community trust is paramount. The next generation of intelligent transportation systems (ITS) must be capable of delivering high-fidelity real-time data without compromising individual rights. Ouster BlueCity with Rev8 directly answers this call.

Ouster BlueCity delivers high-fidelity intelligence coupled with system-level privacy features at the edge

Ouster BlueCity with Rev8 opens up a tremendous amount of new spatial intelligence to traffic engineers and planners. This breakthrough introduces the industry's first native color lidar detection system with up to 500 foot / 150 meter advance detection, simultaneously capturing both color and 3D depth information without the need for complex calibration. This capability provides unmatched situational awareness, delivering precise, high-fidelity digital traffic twins, simplified zone set-up, and more intuitive event analysis through improved color visualization.

Adding native color to lidar detection while addressing concerns about privacy was paramount in Ouster’s product development process. Ouster BlueCity with Rev8 is designed with system-level privacy features at the edge, capturing the movement of multimodal road users in 3D and 2D native color while applying privacy-enhancing techniques to reduce the collection of identifying features such as facial features. As a result, municipalities and DOTs get full situational awareness for the traffic team, while ensuring privacy by design and default for everyone at the crosswalk.


How does Ouster BlueCity provide advanced edge processing?

Ouster BlueCity seamlessly pairs system-level edge-processing power and VRU data blurring at the edge to deliver real-time traffic signal actuation and alerts.

At the core of Ouster BlueCity is our proprietary AI detection software, which utilizes a deep neural network (DNN) to track the trajectories and classifications of multimodal road users. Running on NVIDIA Orin AGX system-on-modules for real-time edge inference, Ouster’s deep learning model is developed entirely in-house and validated against manually annotated lidar data for maximum accuracy.

This secure, edge-processed approach delivers massive advantages over traditional traffic monitoring methods:

  • Advanced AI deep learning: Ouster BlueCity’s proprietary detection model has been trained on 4 million labeled objects collected across 800 sites, encompassing diverse traffic patterns, intersection designs, and environmental conditions.

  • Accurate real-time detection and tracking: Ouster BlueCity’s edge devices process vast amounts of 3D lidar data in real time—delivering low-latency tracking alongside 99% stop bar detection accuracy. This level of precision optimizes actuation, alerts, and analytics to improve signal timings, reduce congestion, and increase safety for all road users.

  • Algorithmic advantages: This approach offers significant advantages over classical algorithms, including improved generalization, greater scalability, reduced computational resource demands, and real-time processing without the need for manual, time-consuming calibration.

Ultimately, Ouster BlueCity proves that municipalities do not have to choose between advanced data and public trust and can access one complete traffic management solution built on the foundation of high performance, accuracy and privacy by design.

To learn more about how Ouster BlueCity, visit ouster.com/bluecity.

Related Articles