Industrial Control Equipment

Monitoring Management System

 

Industrial Control Room System

In the central control room of each intelligent factory, there is a system to monitor the real-time status of each equipment in the factory all at once. There are two types of information need to be collected: the sensors deployed at the operation site to collect production data and analyzed visually for inspection. Another is the monitoring screen directly connected from the equipment end.

Disadvantages of traditional manual work

In the past, when there was no such monitoring system, the operation, maintenance, and abnormality analysis of equipment relied on manual work. Regular inspection was an essential task then, however, manual work was time-consuming and prone to errors, causing managers to spend more time in checking data accuracy. Some factories use an external camera to record employees' actions during the operation instead, nevertheless, the method not only increase the costs, but also make operators to lose trust.

With the rise of IoT, factories have begun to introduce automation to replace part of the manpower. In the process of transformation, centralized management of all equipment in control room requires a system to act as a bridge between old and new equipment. Only when the connection between old and new devices is built that the factory can be successfully upgraded and transformed, and the first challenge will be met is the different signal sources of the equipment.

YUAN has worked closely with all of the world's leading IPC manufacturers and understands the unique characteristics of their industries. We figure out that managers don't want to use external software for risk concern. They also prefer to handle the process on a local site instead of cloud for real-time production. Furthermore, they must take hundreds of devices with different transmission interfaces simultaneously. 

With years of experience in the industry, YUAN has created a powerful NexVDO SDK to help developers entering industrial control market!

NexVDO SDK Helps Digital Transformation

NexVDO SDK is a multimedia software module developed for professional markets. It integrates all standard communication protocols from the underlying video data collection to remote streaming. NexVDO SDK allows developers to easily split the screen to complete front-end data collection, it also provide efficient compression and multi-stream recording to help digitize hundreds of devices. Through only four simple APIs, the SDK can integrate the video of device.

Another advantage of unified recording is that it is easy to identify human errors or system problems when abnormalities occur. Once the equipment screen is centralized in the central control room, a single person can supervise hundreds of devices. In the event of an abnormality, the personnel can apply remote real-time control through the KVM module of NexVDO SDK without going to the site.

In addition, NexVDO SDK integrates ThumbDraw technology and fast codecs native to various GPUs such as Intel GPU, NVIDIA NVENC, AMD VCE, etc. Developers can do hundreds of video renderings and recordings without any performance problems.

Moving from an intelligent factory to a human-free factory

In the past, traditional image recognition mainly used image-matching technology, which is heavily influenced by light source and environment, causing a higher error rate. Nowadays, AI deep learning algorithms are becoming more mature. Most factories are starting to use visual recognition to determine information on the screen and combine environmental webcam images for more diverse applications.

In the recognition module of NexVDO SDK, we support dozens of video AI analyses, which can cover 85% of the factory's application areas. The functions can be mainly divided into three major categories: image content analysis, biometric identification, and behavior analysis. 

The following are user cases:
Type 1. Object Recognition

Identifying defective objects usually requires comparison with the golden sample, so users have to adjust many parameters for testing. Nowadays, developers only input the images in advance to generate AI models. Regardless of object counting, angle, size recognition, or defect detection, the system can completely detect the defects in milliseconds. AI with video integration not only improves inspection accuracy but also saves a lot of human time and successfully improves work efficiency!

Type 2. Text Recognition

Many graphic values or analog images are displayed on the equipment's screen (such as valve and pointer). In the past, operators have to transcribe the data manually. NexVDO SDK can convert the image to text or set rules for automatic feedback, helping old equipment to become intelligent.

Type 3. Behavior Analysis

Managers can use NexVDO SDK's behavior analysis module to monitor industrial safety. It could distinguish whether a person is crossing the electronic fence or identify undesirable crowd gathering and provide early warning. Lastly, we share a case study of a customer who use Smart AI module to classify cups and trays. A tableware cleaning factory in China collected 400 to 500 plates of cups from restaurants. They needed many operators to sort tableware. Since the cups and plates were similar in color and shape, it was difficult to have fixed rules for rubies. They integrated traditional analyzing technology into the system once, but the error rate was too high because of the transparent cup and refraction problem. In classifying objects, it is difficult to recognize the similarity of objects and unique materials without unique algorithms.

NexVDO SDK has a strong AI R&D team that uses an algorithm for glass material to quickly identify many plates, bowls, teacups, and transparent cups of the same color by powerful GPU. The robot arm can grab and group tableware and do the counting simultaneously to speed up process. There is not only machine equipment in the factory but also equipment with video signals, such as robotic arm cameras and environmental safety webcams. A high percentage of the world's significant factories have already implemented automation. The inevitable trend of an intelligent factory is video AI solutions. YUAN can help developers upgrade painlessly in the industrial control field with extensive experience.