WPG Group launches intelligent robotic arm vision system solution based on Intel products

On November 3, 2021, WPG Holdings, a leading semiconductor component distributor dedicated to the Asia-Pacific market, announced that its subsidiary Usun has launched a visual system solution based on Intel NCS2 Neural Compute Stick, suitable for intelligent robotic arms.

Figure 1- A display board of the intelligent robotic arm vision system solution based on Intel products by Dalian Dayoushang

A robotic arm is a machine that can flexibly replace the work of human hands and arms. It is usually an automatic operating device used to grasp, move objects or operate tools according to a fixed program. With the continuous breakthroughs in machine vision, AI and other technologies, the development of intelligent robotic arms is also growing. Dalian Dayou's vision system solution based on Intel NCS2 neural computing stick is applied to robotic arms, which can accurately identify objects and make the grasping, moving and other processes more intelligent.

Figure 2-Application scenario of the intelligent robotic arm vision system solution of DaLian DaYouShang based on Intel products

The core principle of this solution is to use the D435 to identify the color image combined with the deep learning object positioning algorithm YOLOv3 to find the object position (X, Y). When the CPU computing power is insufficient, Intel GPU or VPU (NCS2) can be used to accelerate the calculation, and then the D435 can identify the depth (Z) information of the position corresponding to the depth image, so that the robot arm can correctly reach the correct spatial position (XYZ) to clamp and grab the object.

Figure 3-Block diagram of the intelligent robotic arm vision system solution based on Intel products of Dalian Dayoushang

Intel NCS2 is a VPU (visual processing unit) with a neural compute engine that delivers eight times more performance than the previous generation. Deep learning neural networks such as Caffe, Tensor Flow or MXNet can be directly integrated with the OpenVINO toolkit on NCS2. And these machine learning frameworks are also optimized for the new deep neural network (DNN) inference engine. With a PC and Intel NCS2, developers can quickly start the development of AI and computer vision applications.

NCS2 runs on USB 3.0 and does not require additional hardware, allowing users to seamlessly deploy models trained on their computers to a variety of edge devices without connecting to the Internet or the cloud. Whether developing smart cameras, drones with gesture recognition, industrial robots, or next-generation smart home devices, NCS2 provides developers with the features they need for rapid prototyping.

Core technology advantages:

Intel RealSense D435 camera:

 Low-cost 3D stereo depth camera;

Provide complete SDK for quick integration with the system;

Can quickly scan and provide point cloud information;

Can develop independent innovative functions through ROS integration;

Intelligent 3D object recognition.

 Intel OpenVino Toolkit:

Can optimize the trained model;

Supports training frameworks commonly used in industry and academia;

Can be quickly deployed to Intel hardware platforms such as CPU, GPU, VPU, and FPGA;

Provide commonly used pre-trained models such as SSD, YOLO, etc.;

Provide C++ and Python application examples to shorten the program development cycle.

Intel® Neural Compute Stick 2 (Intel® NCS2):

 Build and scale on the Intel® Movidius™ Myriad™ X vision processing unit (VPU) to deliver exceptional performance per watt per dollar;

Support rapid development on Windows® 10, Ubuntu* or macOS*;

Develop on common frameworks and out-of-the-box sample applications;

Operate without connecting to the cloud;

 Develop prototypes on low-cost edge devices such as Raspberry Pi*3 and other ARM* host devices.

Program Specifications:

 3D camera: Intel RealSense D435 Camera;

Operating system: Windows 10;

Intel NUC Rugged Board with Core i3/i5 Processor;

 AI inference kit: Intel OpenVino Toolkit 2020.03;

Memory : 4GB or more;

Transmission interface: USB 3.0;

Neural Compute Stick: Intel® Neural Compute Stick 2 (Intel® NCS2).

About WPG Holdings:

WPG Holdings is the world's largest semiconductor component distributor and the largest in Asia Pacific*. Headquartered in Taipei, it owns World Peace, Pinjia, Quan Ding and Youshang, with approximately 5,000 employees, over 250 product suppliers, and approximately 100 distribution locations worldwide. Its 2020 turnover reached US$20.65 billion. WPG has created an industry holding platform, focusing on international operation scale and local flexibility. It has long been deeply involved in the Asia Pacific market. With the vision of "Industry First Choice and Channel Benchmark", it fully implements the core values ​​of "Team, Integrity, Professionalism, and Efficiency", and has won the "Global Distributor Excellence Award" for 20 consecutive years. Faced with new manufacturing trends, WPG strives to transform into a data-driven enterprise, build an online digital platform - "Dada.com", and advocate the intelligent logistics service (LaaS, Logistics as a Service) model to help customers face the challenges of intelligent manufacturing together. WPG starts with good intentions and builds trust with technology, hoping to build a competitive ecosystem with the industry, and actively promote digital transformation with the 16-character motto of "focusing on customers, empowering technology, coordinating ecosystems, and creating an era together". (*Market rankings are based on data released by Gartner)