Drone manufacturer DJI has announced the launch of the Manifold, a high-performance embedded Linux computer, with the bold statement that it will enable developers to transform aerial platforms into intelligent flying robots that can perform complex computing tasks and advanced image processing literally on the fly.
With a size of only 11 cm x 11 cm x 2.6 cm and a weight of less than 200 g, the Manifold is optimized for DJI developer platforms and can be mounted to the expansion bay of the DJI Matrice 100 quadcopter.
Designed for developers, the Manifold's built-in Ubuntu operating system supports CUDA, OpenCV and ROS for the research and development of applications. The Manifold can natively run the DJI Onboard software development kit (SDK), access flight data and perform intelligent control and data analysis.
The Manifold is based around the NVIDIA Tegra K1 mobile processor which combines a 4-Plus-1 Quad-core ARM Cortex-A15 CPU and 192 GPU CUDA cores with clock speeds of up to 2.2 GHz and support for DirectX and OpenGL.
Equipped with USB, Ethernet, Mini-PCIe, HDMI, UART, SPI and I2C ports, the Manifold can be connected to a wide array of sensors, monitors and other peripheral devices. The Manifold also has customized ports to connect to the Matrice 100 so that it can process the images captured by the camera mounted to the platform.
DJI claim that the Manifold has the processing power to process high resolution images in real time and for use in artificial intelligence applications such as computer vision and deep learning. This creates the potential for applications which enable a drone to identify and respond to objects in the surrounding environment.
The Manifold is optimized for power efficiency with a peak power consumption of approximately 15W, one fourth of that of a typical laptop according to DJI. The Manifold's Tegra CPU uses its four A15 cores for heavy computing tasks but only a single battery-saver core when performing simple calculations. The number of active cores can be dynamically adjusted to decrease power consumption while the extra core will automatically be activated when needed.
More information can be found on DJI's website.