In the era of Internet of Everything, the application of digital technology in manufacturing has long been an irreversible trend. Although this transformation is a gradual process, at present, the speed of speeding up the transformation has already brought tremendous pressure on enterprises. Previously, HP Enterprise conducted an in-depth investigation with the Industry of Things World Conference. Through this survey, we are looking at how much progress has been made in the industrial IoT projects launched by many companies in the past 12 months. The survey found that only 53% of respondents believe that their industrial IoT projects meet or exceed the expected goals; the remaining 47% said their goals have not been achieved.
As we all know, a company can easily enter the industrial Internet of Things field or easily complete the digital transformation by purchasing technology. Industrial IoT requires a complete architecture in which accessibility can be achieved between the various nodes contained within an organization. All of this also requires a common standard within it and a new technical architecture to create a convergence of IT and OT.
Of course, we are currently facing a key and common problem, that is, the lack of understanding of device connectivity. Because of this problem, even after all the terminal devices are connected to the network, more problems will follow. For example, after the device is connected to the Internet, we don't have a simple and usable tool to manage these devices. Usually there is no way to extract data and information from one language and express it in another. It can be said that there is such a gap between the current programmable logic controller (PLC) data transmission and conversion into an enterprise resource planning (ERP) system. This issue is only a small part of the problems that enterprises will encounter during the digital transformation process.
The purpose of this article is to let the reader know the five main points of the connection puzzle, and propose two solutions:
Device networking, this is a difficult task
When the problem first appeared, many manufacturers in the industrial Internet of Things sector tend to turn these issues into consideration. Once the manufacturer decides to take the risk, these companies suddenly realize that they plan to connect to all the different devices, which involve traditional, modern terminals, closed and open source software, so the plan is implemented. Very difficult, worse, led to serious delays that eventually disrupted the initial schedule.
If you've ever designed a system for a factory, you'll know that connecting and integrating a variety of applications into these factory systems is a nightmare. The consequence of this is that even a simple data collection task can lead to system paralysis, which ultimately takes weeks to repair.
Increasing complexity in the workshop
We know that without a single connection technology, everything can be connected together. As time goes by and technology evolves, industrial workshops are also evolving. Advances in the technology used in the workshop also mean greater complexity within the shop floor. This complexity does not disappear and may even increase over time. As a result, the factory floor mixes a variety of equipment brands with different support agreements and different proprietary data sets.
Therefore, enterprises need to embrace the increasing complexity of the workshop, which means that enterprises accept many mobile parts and terminals in the industrial IoT solution. Only by connecting these components and terminals together can we achieve even greater gains. For business workers, in order to better manage such solutions, they need more professional knowledge. These staff should not treat it as a complex patching system.
System delay issues are highlighted
In order to better solve the complexity of the system, enterprises may choose open platform communication (OPC). OPC is designed to provide a standard network protocol for industrial automation that requires polling to receive data from devices. Polling refers to the location at which the system must ask the device for data at a preset rate, such as once per second or every half hour.
OPC requires multiple steps to send data, not simply from point A to point B. A typical path is as follows: From the PLC to the OPC server to the OPC client, then the OPC client sends it to the local server or cloud network for use and processing.
In order to be further added to the multi-level process, the PLC must be connected separately from any other thing or software application. The connection must input transaction 1 from PLC 1, PLC 1 input transaction 2, and so on. Then transfer the PLC data to the ERP software: the connection of PLC 1 to ERP software 1, ERP software 2, etc. requires code.
There is no doubt that adding polling to all of these layers will result in huge delays.
The data collected is not always accurate
OPC does not provide the ability to guarantee data accuracy and requires extra work to achieve its goals.
Take a Florida-based pharmaceutical company as an example, the company is facing such a challenge. OPC is used to poll devices, which typically receive 3,000 packets per production run, and the system cannot automatically verify which packets are the correct match for the finished batch. To solve this problem, the company's engineering team wrote a large number of complex custom code to double check the data source and receiver to ensure data matching.
Information exchange between traditional equipment and modern equipment
Message Queuing Telemetry Transmission (MQTT) is fast becoming one of the best protocol options for industrial IoT. Many terminal devices currently support the MQTT protocol. The only way to adopt the MQTT protocol is to purchase devices that support the MQTT protocol, but it is conceivable that no one is willing to phase out legacy devices that have been in use for 20 or 30 years and can continue to be used in order to obtain devices that support the MQTT protocol. .
Only when companies want to add new devices, such as the hottest and latest sensors on the market, companies are likely to consider buying sensors based on the MQTT protocol. As a result, companies will not be allowed to invest extra work to enable MQTT equipment to work with its original legacy equipment. For a company with tens of thousands of devices, it may only have 10 devices that support the MQTT protocol, and migrating all devices to MQTT will be a very slow process and may not necessarily solve the networked devices. And legacy issues with new device networking. This means that the system needs more custom coding.
However, in the face of numerous problems, companies are not doing nothing. Two solutions are presented below:
Avoid custom code that uses data-centric IIoT software to map devices directly to applications (or other devices)
This may seem simple, but you may have realized that things are not plug and play as you would expect.
Many IIoT platforms focus on analytics, but because of the inability of their connected devices to quickly collect data, these platforms are seriously lacking in data. Of course, these analysis-focused IIoT platforms are still an analysis platform, but if the data is inaccurate, the analysis results delivered by the platform are actually invalid.
In order to solve this problem, you need to map devices such as PLC directly to the application. The data-centric IIoT software platform is designed for this purpose. Regardless of the communication protocol, it combines legacy and modern devices and provides a central data pipeline for all networked devices and applications, giving organizations complete control over how, when, and where data is being used.
Local drivers - beyond API, OPC and MQTT
Don't be attracted by the idea that application programming interfaces (APIs) and standard protocols give you flexibility, which is more like advertising for APIs, OPCs, and MQTTs.
Every IIoT platform has standard tools for API, OPC and MQTT, but they usually don't have many native drivers. A data-centric platform will have a large number of native drivers that avoid engineers writing custom code inside the enterprise. Thanks to this, the company's terminal equipment can be improved in a few days without spending a few months.
Ningbo Autrends International Trade Co.,Ltd. , https://www.mosvapor.com