Bosch Rexroth is developing a modular automation architecture for agricultural machinery, combining high-performance computing, sensor fusion and position sensing for automated vehicle functions. www.boschrexroth.com The Offroad Robotic Controller ORC2 is designed as the computing core for partial and advanced automation in tractors, front-end loaders and other agricultural machines. It processes data from LIDAR, radar, cameras and other sensors. ORC2 uses the QNX real-time operating system with an integrated ROS2 stack and AI accelerators. Its communication interfaces allow OEMs to integrate different control units and develop their own software functions. Planned software functions include environment and position detection, image recognition, perception logic and bird's-eye-view processing. Bosch Rexroth is also developing 360-degree people detection for monitoring the machine's working area. Sensor-based collision avoidance The Collision Avoidance System combines radar, ultrasonic sensors and intelligent cameras. Radar can detect up to 40 objects simultaneously, while ultrasonic sensing supports up to 12 sensors for near-field monitoring. Intelligent cameras can identify people and distinguish them from other objects. When integrated with the vehicle drive system, the system can support automatic deceleration and emergency braking where specified by the machine manufacturer. The software is developed in accordance with EN ISO 13849, DIN EN ISO 25119 and DIN EN ISO 19014. Communication uses an interface conforming to ISO 21815-2. Position sensing and payload estimation Kinematic Position Sensing uses inertial sensors to determine machine movements. Additional linear or angular sensors can be integrated to provide position information for automated functions such as lift-height limitation. When combined with pressure sensors, the Payload Estimation Loader function can determine the weight of lifted loads across the machine's operating range, including during movement. The system can operate at different component tilt angles and installation positions. Modular software for agricultural machines The BODAS ecosystem allows manufacturers to activate, scale or retrofit software functions according to individual vehicle configurations. Existing software libraries can also support the development of new agricultural machine concepts. The architecture is being prepared for requirements associated with the European Machinery Regulation and the Cyber Resilience Act. Bosch Rexroth states that its portfolio is intended to address the Machinery Regulation requirements applicable from 2027 and Cyber Resilience Act requirements applying from December 2027. Additional Context: This section details technical specifications and competitive benchmarking not included in the original product announcement Bosch Rexroth has not disclosed detailed ORC2 hardware specifications such as AI performance in TOPS, memory capacity, processor configuration or power consumption. This prevents a direct numerical comparison with other edge-AI computing platforms. One established reference is the NVIDIA Jetson AGX Orin platform, which provides up to 275 TOPS of AI performance with a configurable 15 W to 60 W power range. The industrial version provides up to 248 TOPS and supports a 15 W to 75 W range. It uses a 12-core Arm Cortex-A78AE CPU alongside GPU-based AI acceleration. NVIDIA Jetson AGX Orin specifications The comparison should not be interpreted as a performance ranking. Jetson AGX Orin provides published processor-level benchmarks, while ORC2 is presented as an off-highway automation controller combining real-time computing, ROS2, functional-safety considerations, sensor processing and integration with vehicle and hydraulic systems. For agricultural machinery, AI throughput is therefore only one evaluation criterion. Deterministic processing, functional safety, sensor integration and communication with machine-control systems are also important when selecting an automation platform. www.boschrexroth.com Powered by Induportals Media Publishing
