HSINCHU, Nov. 25, 2025 /PRNewswire/ -- As a proven provider of complete Vision-AI solutions for the automotive and robotics markets, oToBrite will unveil a portfolio of next-generation visual perception technologies at iREX 2025. The showcase features the automotive-grade GMSL2 camera series, the VIO (Visual-Inertial Odometry) camera, the oToSLAM multi-camera Vision-AI positioning system, and the new GMSL™ Camera Jetson Developer Kit—demonstrating strong capabilities in automotive imaging, sensor fusion, and AI-driven perception for outdoor robotics, UGVs, humanoid robots, and next-generation smart mobility applications.

oToBrite will introduce its developer kits designed specifically for the Jetson AGX Orin™ and Jetson Orin™ Nano. These solutions accelerate multi-camera vision development on Jetson platforms. The oToAdapter-GMSL-AGX-Orin supports up to eight GMSL2 cameras with 275 TOPS of AI performance, while the oToAdapter-GMSL-Orin-Nano supports up to four cameras with 67 TOPS—both offering plug-and-play integration and full compatibility with all oToBrite GMSL™ cameras.

The new oToCAM274ISP and oToCAM276ISP camera series integrate leading automotive sensors - IMX623 and AR0823 - together with an onboard indie iND88002 camera video processor (CVP). The processor delivers 1,400 MP/s ISP throughput, sub-1 ms latency, 144 dB HDR, and supports up to four cameras with eight real-time video streams. Combined with automotive-grade ISP tuning, the cameras deliver clear, stable, AI-ready imaging under strong backlighting, low illumination, fog, and other challenging outdoor conditions - making them ideal for unmanned vehicles and advanced robotic vision systems.

oToBrite's automotive-grade VIO camera integrates an onboard Inertial Measurement Unit (IMU) - combining an accelerometer and a gyroscope - with Kalman Filter-based fusion to counter vibration-induced image shifts. It provides smooth and stable image output across low-light, high-vibration, and uneven-terrain environments, making it suitable for outdoor unmanned vehicles, AMRs, port systems, and mining equipment.

The oToSLAM multi-camera Vision-AI positioning system - built on SLAM (Simultaneous Localization and Mapping) - offers centimeter-level 3D mapping and localization without HD maps or LiDAR. Supporting both indoor and outdoor environments, it integrates object classification and drivable-area segmentation. Through multi-camera fusion, it provides full 360-degree perception with 1 cm accuracy, ensuring stable performance even in dynamic or complex scenarios.

All these advanced Vision-AI technologies will be showcased at iREX 2025, Booth E7-25, attracting strong interest from global leaders in robotics and intelligent mobility.