Today, when UAVs are accelerating their evolution from "able to fly" to "capable of flying", computing power is productivity. In order to meet the needs of high computing power tasks and cutting-edge algorithm deployment, we launched the SU17-Orin option package, equipped with the NVIDIA Jetson Orin NX core module, giving SU17 UAV powerful edge computing capabilities. It can not only support functions such as independent obstacle avoidance and path planning in complex environments, but also provide a solid foundation for real-machine verification of high-order intelligent algorithms.

01 Core advantages

More computing power

• Equipped with Jetson Orin NX, with a computing power of up to 70TOPS, it can efficiently handle tasks such as image recognition, target tracking, and deep learning inference.

• Equipped with an 8-core Arm Cortex-A78AE CPU and a 1024-core Ampere architecture GPU (32 Tensor Cores), it provides strong support for multi-thread scheduling and deep computing.

Real-time obstacle avoidance

• Seamlessly integrated with the SU17 native quad-camera VIO visual localization system, it achieves centimeter-level depth obstacle avoidance based on the EGO-Swarm algorithm.

• Whether it is a dense indoor environment or complex outdoor terrain, stable obstacle avoidance and path adjustment can be achieved to ensure flight safety.

plug and play

• By connecting the 6PIN cable to the ETH+VCC interface, you can quickly access the SU17 system without additional debugging.

• The overall weight is only 130g, and the lightweight design hardly affects flight performance.

Developer friendly

• Built-in Ubuntu 20.04 system, pre-installed with common tools such as ROS Noetic and OpenCV 4.7.0.

• Perfectly compatible with the PrometheusUAVopen-source framework, it supports the rapid construction of a complete intelligent flight process from perception to control.

02 Technical configuration

This optional package is deeply integrated with the Prometheus UAV software framework and has built-in complete path planning and autonomous flight solutions. Combined with the EGO-Swarm algorithm, intelligent obstacle avoidance and global optimal path generation can be achieved, allowing developers to quickly build a complete intelligent flight system covering environment perception, decision planning, and motion control.

Applicable scenarios

• Universities and scientific research institutions: high computing power tasks and cutting-edge algorithm deployment;

• High-precision 3D modeling and real-time perception;

• Autonomous flight in complex environments, etc.

03 Preview

Adapted to FAST-LIVO2

The SU17 platform will integrate the FAST-LIVO2 algorithm developed by the MaRS Laboratory of the University of Hong Kong to achieve a higher-precision and more robust SLAM system through the tight coupling of LiDAR, IMU and vision. The computing power and graphics memory of SU17-Orin can significantly improve the real-time performance and visualization quality of mapping, and provide full-link support for complex tasks:

• Real-time rendering of mapping effects improves the operator's understanding of the environment;

• Share the load of path planning and perception modules to enhance the real-time response of the SLAM core.

Product wiki:https://monojson.com/s/AsGBT