Luxonis has unveiled its OAK 4 devices, a next-generation edge AI camera system that brings 52 TOPS of on-device inference—a 40x increase over the previous generation. Designed for robotics, industrial automation, smart cities, retail, logistics, and advanced driver-assistance systems (ADAS), OAK 4 enables machines to perceive the world without relying on external cloud compute.
“After years of sustained engineering investment, OAK 4 represents the most advanced generation of our edge-AI platform to date,” said Bradley Dillon, CEO of Luxonis. “Deployments have made it clear: developers need higher-throughput compute, expanded multimodal sensing, and rugged hardware. OAK 4 delivers on all fronts.”
Industry-Leading Compute and Sensor Fusion
Each OAK 4 device integrates a Qualcomm® QCS8550 processor—a heterogeneous platform with ISP, GPU, and Neural Signal Processing Unit (NSPU)—running Yocto Linux OS. The devices are fully standalone, supporting multiple computer vision models at high frame rates, including Luxonis’ proprietary Neural Stereo Depth models.
Key hardware features include:
- Stereo depth camera pair and 48MP RGB camera
- IR dot projector and IR illumination LED
- 9-axis IMU and microphone
- Wide FOV for navigation
- IP67-rated industrial chassis for shock, vibration, and harsh conditions
- USB or PoE power with M8 connector for I/O connectivity
Patent-pending Dynamic Calibration ensures precise stereo depth even under changing environmental conditions.
“By combining rugged, high-performance hardware with a unified cloud platform, Luxonis is addressing one of the biggest challenges in deploying AI at scale,” said Anshuman Saxena, VP & GM, ADAS & Robotics, Qualcomm Technologies. “This empowers developers to build smarter, more resilient systems with unprecedented precision.”
Luxonis Hub: Bridging Hardware and Cloud
In parallel with OAK 4, Luxonis is launching Luxonis Hub, a cloud-based platform that simplifies device deployment, monitoring, and over-the-air updates. Hub mitigates model drift with ‘Snaps,’ which collect edge data to improve AI models automatically. What once required specialized engineering can now be accomplished in just 5 minutes, even with limited technical expertise.
Applications and Developer Impact
With over 4.5 million SDK downloads, Luxonis’ platform supports a range of applications from autonomous mobile robots (AMRs) to IoT sensors. Its SDK abstracts complex vision tasks—sensor control, stereo depth, neural inference, and hardware acceleration—into simple, composable APIs, allowing developers to deploy reliable real-time vision pipelines with flexibility for customization and scaling.
OAK 4 and Luxonis Hub together represent a major step forward for edge AI, combining industry-leading compute, sensor fusion, and cloud integration in a fully self-contained system that reduces costs and complexity while enabling high-performance AI at the edge.
Power Tomorrow’s Intelligence — Build It with TechEdgeAI
