Sundance Multiprocessor Technology has launched the VCS-1 — a new embedded platform explicitly designed for precision robotics, which the company states incorporates real-time vision, control, and sensor applications.
“The VCS-1 embedded processor module is optimized for computer vision, edge AI and deep learning requirements. It weighs just 300g, has a low power consumption of typically 15W, and is highly compatible with a wide range of commercially available sensors and actuators.”
The VCS-1 utilizes the PC/104 form factor, providing industry-standard compatibility and expandability, and is composed of two main boards. The EMC2 board hosts a Xilinx Zynq MPSoC, which packs a quad-core Arm Cortex-A53 or Cortex-A9 depending on preference. It also features an Arm Mali-400 GPU, and a Cortex-R5 RPU (real-time processing unit) that are tasked for managing real-time events and programmable FPGA logic for hardware acceleration of AI apps that are employed by a Trenz TE0820 SoM.
The Sundance External Interface Card is loaded with a ton of I/O and connection options, including multiple USB 3.0 ports for connecting cameras and sensors, such as Intel’s RealSense T625 tracking camera and Stereo Labs’ Zed depth cameras. The board is even compatible with Arduino and Raspberry Pi sensor and actuator modules, and can mimic a PC by incorporating HDMI, SATA, and Ethernet networking ports. ADC, DAC, GPIO headers for expansion are featured as well, allowing users to connect nearly anything to the board.
On the software end, the VCS-1 can take advantage of ROS, OpenCV, Xilinx SDSoC, MQTT, Ubuntu, Python, and TULIPP’s STHEM toolchain. More information on Sundance’s VSC-1 embedded robotic platform can be found on the company’s website.