The Nova Orin Autonomous Mobility Robot (AMR) reference platform has been enhanced by NVIDIA to speed up development
What are the recent updates of the Nova Orin Autonomous Mobility Robot?
The Nova Orin autonomous mobility robot (AMR) reference platform has been enhanced by NVIDIA to speed up development. Three reference platform configurations’ specifications are made accessible. Two rely on a single Jetson AGX Orin, while one makes use of two Orin modules, to run the NVIDIA Isaac robotics stack and the Robot Operating System (ROS) with the GPU-accelerated framework.
The Nova Orin platform aims to reduce costs and increase dependability in the international development and deployment of AMRs.
Autonomous mobile resources, or AMRs, operate in ad hoc settings. They can maneuver around obstacles and don’t require fixed, preprogrammed courses. They are excellent for logistical chores like cleaning, giving hospitality, providing roving security, and making last-mile deliveries since they can move objects in factories, warehouses, and distribution centers.
AMR producers have been creating robots for a long time by obtaining and integrating computational gear, software, and sensors internally. Costs associated with this time-consuming method include lengthy go-to-market pipelines, years of engineering resources, and a distraction from developing domain-specific products.
Creating a Very Flexible and Powerful Reference Architecture
The Nova Orin reference architectural designs come with specific use cases. The other Orin-based design has a safety programmable logic controller in addition to safety-certified sensors, while the first does not. The third architecture uses a dual Orin-based design with visual AI to enable functional safety.
Support is offered for inertial measurement systems, lidars, stereo cameras, and ultrasonic sensors. The sensors selected for industrial applications strike a balance between dependability, affordability, and performance. AMRs that are secure and cooperative can be developed and deployed with the help of the sensor suite, which provides the multimodal diversity of coverage required.
The stereo cameras and fisheye cameras were developed by NVIDIA in partnership with camera partners. All sensors are calibrated and time-synchronized, and they all come with drivers for consistent data collecting. AMRs can identify objects and barriers in a range of contexts thanks to these sensors’ simultaneous localization and mapping capabilities (SLAM).
Constructing a complete AI platform for OEMs and ISVs
The base OS includes design documents for integrating it with robots as well as firmware, drivers, and adapter tools for all the hardware. It is simple to connect a ROS-based robot application to Nova.
The sensors will have validated models in Isaac Sim for application development and testing without using a real robot.
The cloud-native data collection tools take care of the time-consuming task of building data pipelines for the massive amount of sensor data needed for training models, troubleshooting, and analytics. Modern GEMs developed for Nova sensors are GPU accelerated with the assistance of the Jetson Orin platform, providing key building blocks such as visual SLAM, stereo depth estimation, obstacle recognition, 3D reconstruction, semantic segmentation, and posture estimation.
The need for quick development of city-scale, high-fidelity 3D maps for indoor environments is another issue that Nova Orin addresses. These produced maps can be used for simulation, fleet planning, and robot navigation. The maps can be regularly updated using the robots’ data as well.
Industry ready AMRs
As robotics systems advance, future AMRs must enable the deployment and management of the essential AI software installed.
Nova Orin supports secure over-the-air upgrades, device management, and monitoring to simplify deployment and keep maintenance costs down. The platform’s open, modular design allows developers to use any or all of its features and alter them, which enables the creation of robotics applications quickly.
In an effort to further lower the cost and increase the dependability of AMRs, NVIDIA is directly collaborating with regulatory agencies to develop safety technology that can be used with vision. Additionally, we offer a navigation software development kit to aid programmers in creating applications fast.
AMRs must collaborate safely and effectively in large-scale operations to increase productivity in factories and warehouses. This transition will be fueled by high levels of autonomy by Nova Orin’s 3D sensing.
