Stereolabs has introduced TERRA AI, the latest addition to the company’s vision platform portfolio. This new technology is a groundbreaking, general-purpose AI, enabling robots and autonomous systems to operate with heightened safety and exceptional precision.
Designed for an optimized implementation of front, back, and surround cameras for autonomous robots, TERRA AI leverages monocular and stereo vision to provide 4D volumetric perception with unprecedented accuracy.
This new technology performs depth, localization, semantic, obstacle detection, and fusion from multiple cameras in both space and time. It represents a leap forward in the ability for autonomous robotics to perceive space.
Vision-based autonomy powered by high-performance computing
Vision-based full autonomy for robots is a non-trivial task that requires multiple features to be performed simultaneously and efficiently on embedded hardware: depth estimation for spatial awareness, visual-inertial odometry for indoor and outdoor localization, semantic understanding for obstacle classification, terrain mapping and robust detection of people surrounding the robot for safety.
TERRA AI is a groundbreaking large Neural Network Architecture crafted with fast, memory-efficient, low-level code to perform those tasks simultaneously on low-power embedded computers, enabling the processing of high-frequency, high-volume data from ZED cameras on embedded NVIDIA® Jetson™ platform.
Combining the high-performance NVIDIA® Jetson™ module with TERRA AI, the Stereolabs vision platform is engineered with a blend of industry-proven hardware – the ZED cameras – and software solutions.
Stereolabs new mono camera, the ZED X One, is available in 4K and global shutter and is designed for multiple-camera surround vision or large baseline stereo applications. Regarding the software, Stereolabs supports various functions needed for enhanced perception around machines to allow for safe navigation and manipulation.
Boosting productivity among resource constraints and labor shortages
Many industries – from agriculture to logistics to construction and beyond – must produce at a higher volume and execute faster than ever before, all with fewer resources. At the same time, they’re contending with labor shortages and what can be hazardous working conditions. Autonomous robots are key to solving these challenges, but they need the ability to evaluate their surroundings, interpret data, and complete tasks independently to be effective. Advanced sensing technology is crucial for unlocking those capabilities, and TERRA AI is answering the call.
TERRA AI offers a scalable way to deploy autonomous machines – such as tractors, mowers, utility vehicles, AMR, robotic arms and more – by leveraging low-cost cameras combined with NVIDIA® Jetson™ Orin GPU to provide a panoramic 360° surround view around the robot. The elimination of LiDAR or radar indicates a significant shift towards lower-cost scalable solutions.
Cecile Schmollgruber, CEO at Stereolabs explains; “As we further collaborate with manufacturers and system integrators, our latest TERRA AI technology aims to address a growing need for a more open, adaptable and scalable vision platform delivering 360 degree perception and driving true autonomy. Terra AI will allow companies across industries to invest in autonomous robotics at scale, helping them run their operations with greater efficiency and precision at lower cost.”
TERRA AI is now in research preview. It will be available to Stereolabs Enterprise customers in Q2 2024, and later on to all users via the ZED SDK.
The ZED SDK is an advanced, open and scalable AI Vision platform, used by 100,000+ developers around the world. It is designed to help makers and industry leaders produce safe, smart autonomous applications running 24/7.
Find developers of Artificial Intelligence Technology for autonomous systems >>
