Announcing the Foxglove SDK, a unified toolkit that simplifies integrating your robotics stack with Foxglove. Read the full announcement →
Announcing the Foxglove SDK, a unified toolkit that simplifies integrating your robotics stack with Foxglove. Read the full announcement →
In the rapidly advancing fields of robotics and artificial intelligence, perception is everything. How a robot interprets and navigates its environment determines its capabilities, reliability, and autonomy. The growing complexity of real-world environments demands something more powerful than traditional 2D cameras: 3D vision.
2D cameras provide images with width and height — but lack depth, making them inherently limited for tasks that require spatial understanding. Robots using 2D vision can struggle to estimate distances, differentiate between overlapping objects, or adapt to dynamic and cluttered environments. This shortcoming hinders reliable object avoidance, mapping, and interaction.
Depth perception enables robots to localize themselves in space, identify obstacles, and plan collision-free paths — all in real time. Whether it’s an autonomous vehicle identifying pedestrians or a drone navigating through a dense forest, the ability to perceive the third dimension is crucial for safety, precision, and performance.
StereoLabs ZED cameras use stereo vision to reconstruct the environment in 3D — much like human eyes. By capturing synchronized images from two lenses and applying advanced stereo matching algorithms, ZED generates high-resolution depth maps and 3D point clouds in real time.
These capabilities are tightly integrated with Simultaneous Localization and Mapping (SLAM), allowing robots to build maps of unknown environments while keeping track of their own position. ZED’s visual-inertial odometry blends camera input with inertial measurements for accurate, drift-resistant tracking.
ZED’s onboard AI models enable real-time object detection, tracking, and semantic segmentation — making it possible to differentiate between cars, humans, and other objects. This spatial awareness fuels intelligent decision-making and context-aware interaction, giving robots the ability to “understand” their surroundings rather than just react to them.
Understanding what your robot “sees” and “senses” is essential — especially when something goes wrong. That’s where Foxglove comes in. With the new integration between ZED Cameras and Foxglove, you can now stream and visualize ZED stereo video, depth maps, and motion data in real time.
You can correlate 3D camera data with sensor telemetry and system diagnostics, dramatically accelerating your debugging and development workflows. Instead of parsing log files, you can see and fix issues visually — in a few clicks.
By integrating ZED with Foxglove, you can:
This empowers you to iterate faster, improve autonomy stacks, and push your perception systems toward real-world deployment readiness, even faster.
From delivery robots to warehouse automation, 3D vision is critical. ZED cameras allow mobile robots to avoid obstacles, localize within dense spaces, and map surroundings with high accuracy.
Drones face unique challenges like high-speed navigation and rapid environmental change. ZED’s long-range stereo vision allows drones to detect obstacles several meters away and execute smooth, safe landings using 3D terrain data. Debugging flight perception through Foxglove helps eliminate guesswork and reduces mission failures.
For autonomous cars, spatial awareness isn’t optional — it’s foundational. ZED’s real-time depth sensing and semantic understanding provide robust perception, while Foxglove enables inspection of every frame, object, and depth map during testing and validation.
The synergy between deep learning and depth data is unlocking new capabilities in object recognition, behavior prediction, and human-robot interaction. ZED cameras provide rich training data for AI models, while Foxglove’s visualization tools help inspect, validate, and fine-tune model outputs during development.
3D vision goes beyond robotics. In augmented reality, ZED enables real-world occlusion and interaction. In smart cities, depth cameras can monitor traffic, crowd flow, and infrastructure in 3D. These systems rely on accurate, real-time spatial data to operate effectively — exactly what StereoLabs provides.
As physical AI systems become more complex, they need better development tools. The integration of StereoLabs and Foxglove represents a major leap forward: enabling seamless multimodal data exploration, faster iteration loops, and a deeper understanding of real-world behavior. It transforms debugging from a bottleneck into a strategic advantage.
Ready to see what your robot sees? Plug in your ZED camera, fire up Foxglove, and explore the future of robotics and physical AI from a new perspective.
