Click the blue text for more exciting information

Remote Monitoring Projects That Were Ruined by Industry 4.0

Honestly, I have been in this industry for almost 20 years and have seen too many so-called “factory informationization” projects end up as a pile of scrap metal. Last summer, a paper mill in Zhejiang spent a lot of money on a so-called “cloud monitoring system,” and what happened?It dropped connections every few days, it was a disaster! A bunch of salespeople keep shouting about Industry 4.0 and digital twins, but I want to ask, has the communication stability of your PLC been resolved yet?

Remote monitoring is simply about transmitting on-site data to the cloud; as long as you can see it, that’s enough. But those consultants who only know how to make PowerPoint presentations insist on making it overly complicated.The fundamental problem is that everyone does not understand the underlying principles, thinking that data collection is just SCADA connecting automatically. Nonsense!

In 2021, I handled a case for a food factory that imported a complete set of Siemens S7-1500 configuration, and the remote monitoring system crashed every few days. The factory manager was frantic, and the IT department was blaming each other. In the end, I went on-site to see what the problem was. What did I find? The communication configuration was all set to default parameters, and regardless of how many devices or how much data there was, they stuffed everything into one OPC. Isn’t it normal for it to crash?

// This is their original code, it makes you want to curse

DATA_BLOCK "RemoteMonitor"

{ S7_Optimized_Access := 'TRUE' }

VERSION : 0.1

VAR

Device_1_Data : Array[0..999] of Word; // Stuffing so much data into one array, are you crazy?

UpdateRate : Time := T#100MS; // Refreshing every 100ms, can the network handle it?

Connection_Status : Bool;

END_VAR

The old Siemens remote communication module is completely different from AB’s. I have always felt that AB does a much better job in this regard; at least they understand data grouping and prioritization.Don’t even think about domestic PLCs in this area; a few years ago, a certain domestic brand boasted about their remote monitoring, but after connecting less than 50 points, they started losing data. Can this still be called industrial control?

Ultimately, the core issues of remote monitoring and interconnectivity are not whether you can connect, but rather solving these problems:

1. Communication stability (this is the foundation; without stable communication, everything else is nonsense)

2. Data grouping priority (can critical data and ordinary data be processed separately?)

3. Disconnection recovery mechanism (many projects completely ignore this; once disconnected, all data is lost)

4. Security (with so many industrial control viruses now, would you dare expose the system directly to the public internet?)

Last year, an automotive parts manufacturer spent a lot of money on an “industrial internet platform,” but encountered a major pitfall – data loss. The reason is simple,they didn’t even do the basic disconnection reconnection properly! When I looked at their code, they didn’t even have a simple cache queue; these software developers really don’t understand automation.

// The part of the code I modified later

#DB100 // Communication buffer

Buffer_Size := 240; // Set buffer size according to actual conditions

Connection_Retry := TRUE; // Enable automatic reconnection

Store_On_Disconnect := TRUE; // Local storage during disconnection

// Added a simple check; when disconnected, write data to local SD card

IF #Connection_Status = FALSE THEN

CALL "Write_To_Local_Storage"(

Data := #Current_Data,

Timestamp := #System_Time);

END_IF;

By the way, many engineers think OPC UA is a panacea that can solve everything. Nonsense!OPC UA also has problems with high-frequency data collection, especially for millisecond-level high-speed collection; if you don’t do a good job with layered caching design, it will still crash. In 2022, I went to a steel factory in Hebei to troubleshoot; their system claimed to be “fully interconnected,” but they couldn’t even handle the data from one rolling mill, and the system crashed every peak period.

Interconnectivity is not just about setting up a platform and plugging in a network cable.The core is to understand the protocol differences between various PLCs. Siemens’ S7 protocol, AB’s CIP, Mitsubishi’s MC, the underlying implementations are completely different. If you don’t do a good job with protocol conversion and adaptation, what interconnectivity are you talking about!

Finally, let me share some practical experience. If you really want to do PLC remote monitoring:

1. Be cautious in choosing communication protocols; don’t be fooled by those new concepts.Use what you need; stability is the most important

2. Data caching mechanisms must be well established; local storage + disconnection retransmission is basic operation

3. Hierarchical collection; important data should have high priority, while ordinary data can be collected at lower frequencies

4. Security isolation is essential; at the very least, implement one-way data flow protection

The funniest thing is those projects that casually mention “edge computing.” A few months ago, I went to a certain new energy vehicle factory, and their so-called “edge computing box” was just an industrial computer running Linux; it could only compress video and kept crashing.Is this Industry 4.0? Nonsense!

To truly achieve remote monitoring, you must start with basic communication to ensure data integrity and real-time performance. This task sounds simple, but it is full of pitfalls. Most of the so-called “one-click interconnect” solutions on the market are just scams. Don’t be fooled by those flashy interfaces,stability and reliability are the key!!

Oh, and it’s best not to blindly trust those “industry solutions”; test more and accumulate experience. There is no perfect solution, only the one that best fits your factory. I have seen too many projects that start big and comprehensive, but end up spending a lot of money with a ton of problems. It’s better to be pragmatic, take it step by step, and stabilize the small system first before moving on.

Disagreements lead to me watching