Skip to content

Commit 692e0e3

Browse files
emqx-ci-robotRed-Asuka
authored andcommitted
sync blog
1 parent 0a0fa97 commit 692e0e3

5 files changed

+303
-0
lines changed

README-ZH.md

Lines changed: 1 addition & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -325,6 +325,7 @@ EMQX is the world's most scalable open-source MQTT broker with a high performanc
325325
## [EMQX Cloud](https://www.emqx.com/zh/blog/category/cloud)
326326
The easiest way to start MQTT service. Connect your IoT devices to any cloud without the burden of maintaining infrastructure.
327327

328+
- [[计划内维护通知] EMQX Cloud Serverless 平台 6 月 3 日升级通知](https://www.emqx.com/zh/blog/emqx-cloud-serverless-deployment-upgrades-june-2025) ([Edit](https://github.com/emqx/blog/blob/main/zh/202506/emqx-cloud-serverless-deployment-upgrades-june-2025.md))
328329
- [开发者分享:利用 EMQX Cloud 与 ESP32 微控制器实现智能液冷散热系统](https://www.emqx.com/zh/blog/a-developer-s-journey-with-esp32-and-mqtt-broker) ([Edit](https://github.com/emqx/blog/blob/main/zh/202402/a-developer-s-journey-with-esp32-and-mqtt-broker.md))
329330
- [MQTT 平台解读:构建现代物联网应用的必备组件](https://www.emqx.com/zh/blog/mqtt-platform-essential-features-and-use-cases) ([Edit](https://github.com/emqx/blog/blob/main/zh/202402/mqtt-platform-essential-features-and-use-cases.md))
330331
- [EMQX Cloud 和工业物联网:从边端数据到 InfluxDB 3.0 的工业数据集成](https://www.emqx.com/zh/blog/supercharging-iiot-with-mqtt-edge-intelligence-and-influxdb) ([Edit](https://github.com/emqx/blog/blob/main/zh/202311/supercharging-iiot-with-mqtt-edge-intelligence-and-influxdb.md))

README.md

Lines changed: 2 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -322,6 +322,7 @@ EMQX is the world's most scalable open-source MQTT broker with a high performanc
322322
## [EMQX Cloud](https://www.emqx.com/en/blog/category/cloud)
323323
The easiest way to start MQTT service. Connect your IoT devices to any cloud without the burden of maintaining infrastructure.
324324

325+
- [[Scheduled Maintenance Notice] EMQX Cloud Serverless Deployment Upgrades: June 2025](https://www.emqx.com/en/blog/emqx-cloud-serverless-deployment-upgrades-june-2025) ([Edit](https://github.com/emqx/blog/blob/main/en/202506/emqx-cloud-serverless-deployment-upgrades-june-2025.md))
325326
- [Your Next Move: A Strategy for Upgrading from EMQX Standard](https://www.emqx.com/en/blog/your-next-move-a-strategy-for-upgrading-from-emqx-standard) ([Edit](https://github.com/emqx/blog/blob/main/en/202404/your-next-move-a-strategy-for-upgrading-from-emqx-standard.md))
326327
- [Strategic Update: Refocusing EMQX Platform Service Offerings](https://www.emqx.com/en/blog/strategic-update-refocusing-emqx-platform-service-offerings) ([Edit](https://github.com/emqx/blog/blob/main/en/202404/strategic-update-refocusing-emqx-platform-service-offerings.md))
327328
- [Smart Spending: How to Avoid Unexpected Costs with EMQX Serverless](https://www.emqx.com/en/blog/smart-spending-how-to-avoid-unexpected-costs-with-emqx-serverless) ([Edit](https://github.com/emqx/blog/blob/main/en/202403/smart-spending-how-to-avoid-unexpected-costs-with-emqx-serverless.md))
@@ -511,6 +512,7 @@ Build a reliable, efficient and industry-specific Internet of Vehicles platform
511512

512513
## [Industrial IoT | Unified Namespace | Sparkplug](https://www.emqx.com/en/blog/category/industrial-iot)
513514

515+
- [Natural Language Meets Industrial Data: Pioneering AI-Driven IIoT Data Exploration](https://www.emqx.com/en/blog/pioneering-ai-driven-iiot-data-exploration) ([Edit](https://github.com/emqx/blog/blob/main/en/202506/pioneering-ai-driven-iiot-data-exploration.md))
514516
- [Modbus Protocol 2025: From Basics to IIoT Integration](https://www.emqx.com/en/blog/modbus-protocol-the-grandfather-of-iot-communication) ([Edit](https://github.com/emqx/blog/blob/main/en/202503/modbus-protocol-the-grandfather-of-iot-communication.md))
515517
- [ Siemens PLC and MQTT Integration: A Step-by-Step Guide](https://www.emqx.com/en/blog/siemens-plc-and-mqtt-integration) ([Edit](https://github.com/emqx/blog/blob/main/en/202501/siemens-plc-and-mqtt-integration.md))
516518
- [Industrial IoT Systems: Benefits, Essential Capabilities, and Best Practices](https://www.emqx.com/en/blog/industrial-iot-systems) ([Edit](https://github.com/emqx/blog/blob/main/en/202407/industrial-iot-systems.md))
Lines changed: 79 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,79 @@
1+
We're upgrading our global Serverless platform to enhance stability, performance, and observability. **Dedicated and BYOC deployments are not affected.**
2+
3+
## Upgrade Schedule
4+
5+
We'll perform upgrades across three regions to minimize business impact:
6+
7+
**Europe Region**
8+
9+
- **Date:** June 6, 2025
10+
- **Time:** 02:00 - 03:00 UTC
11+
- **Downtime:** 30-60 seconds
12+
13+
**Asia-Pacific Region**
14+
15+
- **Date:** June 10, 2025
16+
- **Time:** 02:00 - 03:00 UTC
17+
- **Downtime:** 30-60 seconds
18+
19+
**North America Region**
20+
21+
- **Date:** June 17, 2025
22+
- **Time:** 02:00 - 03:00 UTC
23+
- **Downtime:** 30-60 seconds
24+
25+
## What to Expect
26+
27+
### During Maintenance
28+
29+
- MQTT connections will drop briefly (2-3 times, 30-60 seconds each)
30+
- Creating or modifying Serverless deployments will be disabled
31+
- Workloads with no active devices are unaffected
32+
33+
### After Maintenance
34+
35+
- Service hostnames remain unchanged
36+
- IP addresses will change
37+
- Update firewall/security group allowlists if needed
38+
39+
## What You Need to Do
40+
41+
### Before Maintenance
42+
43+
- Ensure your MQTT clients support automatic reconnection
44+
- Use domain names instead of IP addresses when possible
45+
46+
### After Maintenance
47+
48+
- Update IP allowlists if applicable
49+
- Verify all devices reconnect successfully
50+
51+
## Regional Time Zones
52+
53+
### Europe Region (June 6)
54+
55+
- London: 03:00 - 04:00 (BST)
56+
- Paris/Berlin: 04:00 - 05:00 (CEST)
57+
- Dubai: 06:00 - 07:00
58+
59+
### Asia-Pacific Region (June 10)
60+
61+
- Singapore/Beijing: 10:00 - 11:00
62+
- Tokyo: 11:00 - 12:00
63+
- Sydney: 12:00 - 13:00
64+
65+
### North America Region (June 17)
66+
67+
- New York: 22:00 - 23:00 (June 16, EDT)
68+
- Los Angeles: 19:00 - 20:00 (June 16, PDT)
69+
- São Paulo: 23:00 - 00:00 (June 16-17, BRT)
70+
71+
## Support
72+
73+
- **Status updates:** https://status.emqxcloud.com
74+
- **Technical support:** Submit a support ticket
75+
- **Questions:** Our SRE team will prioritize your requests
76+
77+
We follow strict procedures to ensure smooth transitions. Thank you for your trust in EMQX Cloud.
78+
79+
**EMQX Cloud SRE Team**
Lines changed: 176 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,176 @@
1+
## **Overview**
2+
3+
With the rapid advancement of artificial intelligence (AI) technologies, industrial enterprises are increasingly exploring their potential in smart manufacturing and industrial data analytics. Meanwhile, the Industrial Internet of Things (IIoT) continues to drive the evolution of device connectivity and data collection, providing a robust data foundation for AI-enabled industrial applications.
4+
5+
This article demonstrates an application for IIoT data exploration and analysis, integrating multiple cutting-edge technologies. The application combines large language models (LLM), Retrieval-Augmented Generation (RAG), Model Context Protocol(MCP), Sparkplug B protocol, MQTT message broker, time-series databases, and relational databases. It delivers an intelligent, interactive industrial data application, enabling users to query device data, analyze historical metrics, and monitor status by simply entering natural language queries in a dialogue box. The system also vectorizes product manuals, automatically providing users with relevant solutions when error codes are reported, facilitating rapid issue identification and resolution.
6+
7+
Readers will gain insights into developing AI-driven IIoT applications and leveraging AI’s value in industrial manufacturing scenarios.
8+
9+
## **Scenario Description**
10+
11+
We assume a typical industrial scenario with the following structure defined using the SparkPlug B protocol:
12+
13+
- **Factory**: Identified as `factory_1` in SparkPlug.
14+
- **Production Line**: Identified as `assembly_1` in SparkPlug.
15+
- **Specific Device**: Identified as `demo` in SparkPlug (nicknamed "Optimus Prime," with this information stored in a third-party relational database).
16+
17+
In this demo, we simulated the runtime data of an ABB FlexPendant virtual device and used its public documentation as the knowledge base. ABB FlexPendant is a handheld touchscreen device used for programming and controlling industrial robots. This device serves as the user interface for the robot controller, allowing operators to perform various operations such as modifying and running programs, teaching new robot actions, and adjusting parameters. To simplify the scenario, we define the following data tags, representing:
18+
19+
- Voltage and current values of the robotic arm.
20+
- Diagnostic error codes.
21+
22+
```json
23+
{
24+
"robotic_arm": { "voltage": 3.14, "current": 5.0},
25+
"diagnose": {"error_code": 50153}
26+
}
27+
```
28+
29+
### **Interactive Scenario Examples**
30+
31+
Users can interact with the system through an AI dialogue box. The system incorporates built-in Agentic capabilities and integrates various MCP services, automatically generating relevant query requests based on user needs and returning data analysis reports. Typical interaction scenarios include:
32+
33+
- **List the tree structure of device** `demo`: The AI queries relevant data and displays the logical structure diagram of the device.
34+
- **Query the latest** `robotic_arm/voltage` tag data for device `demo`: The AI retrieves the most recent data reported for the `robotic_arm/voltage` tag of the `demo` device.
35+
- **Analyze the past week's** `robotic_arm/voltage` tag data for device `demo`: The AI queries and analyzes one week's worth of `robotic_arm/voltage` data for the `demo` device stored in the TDengine database. Given the potentially large data volume, the AI will automatically aggregate the data based on built-in instructions and provide a corresponding analysis report.
36+
- **Query the latest** `robotic_arm/current` tag data for the device nicknamed "Optimus Prime": Since the user queries using the nickname "Optimus Prime," the AI retrieves the corresponding device ID (`demo`) from the MariaDB MCP service and fetches the latest data point.
37+
- **Query the latest** `diagnose/error_code` tag data for device `demo`: The AI retrieves the latest `diagnose/error_code` tag, uses RAG technology to access relevant manual knowledge from the vector database, and presents the user with a comprehensive error report and solution.
38+
- **Analyze the past day's** `diagnose/error_code` tag data for device `demo`: The AI queries the diagnostic error codes from the past day, combines them with operational manual data from the vector database, and generates a comprehensive error report, solutions, and recommendations.
39+
40+
## Demo
41+
42+
### **List the tree structure of device** `demo`
43+
44+
Below is a screenshot of the result after the user inputs "List the tree structure of device `demo`".
45+
46+
![image.png](https://assets.emqx.com/images/3af5aa7a3e63b3fbf4060f5e1e505a2c.png)
47+
48+
### Query the latest `robotic_arm/current` tag data for the device nicknamed "Optimus Prime"
49+
50+
1. **Overview**
51+
- **Analysis Time Range:** Latest data point (real-time)
52+
- **Record Count:** 1 entry
53+
- **Key Findings:** The latest current value for the device with alias **"Optimus Prime"** (actual device key: **demo**) is **2.5A**.
54+
55+
2. **Detailed Analysis**
56+
57+
![image.png](https://assets.emqx.com/images/9f370170ce2075a60357213755653b71.png)
58+
59+
3. **Diagnosis & Recommendations**
60+
- **Current Status Assessment:** The current value is within the normal range (typical operating current for ABB FlexPendant robots is 2–6A).
61+
- **Recommended Actions:**
62+
- Continue monitoring the trend of this value.
63+
- If it consistently stays below 3A, consider checking:
64+
- Whether the robotic arm is underloaded
65+
- Whether the program is executing completely
66+
- Whether the power supply voltage is stable
67+
68+
4. **Additional Notes**
69+
70+
This report is based on a single-point instantaneous data sample. For trend analysis, it is recommended to query data over a time range. Normal current fluctuation for ABB FlexPendant devices is within ±10% of the nominal value.
71+
72+
### Analyze the past week's `robotic_arm/voltage` tag data for device `demo`:
73+
74+
**Device Name:** demo
75+
76+
**Tag:** robotic_arm/voltage
77+
78+
**Time Range:** 2025-05-13 to 2025-05-20
79+
80+
1. **Data Overview**
81+
- **Total Raw Records:** 8,982
82+
- **Sampling Strategy:** 6-hour average aggregation (28 data points in total)
83+
- **Voltage Range:** -17.40V to 3.40V
84+
85+
2. **Key Findings**
86+
- **Stable Main Voltage:** 3.40V (observed during most time periods)
87+
- **Abnormally Low Voltage:** -17.40V (requires further investigation)
88+
89+
90+
91+
### Query Latest `diagnose/error_code` Data for Device `demo` and Retrieve Related Knowledge Base Information
92+
93+
1. **Overview**
94+
- **Device Name:** demo
95+
- **Analysis Time:** Latest data
96+
- **Tag:** diagnose/error_code
97+
- **Key Finding:** The device has reported **error code 10139**
98+
99+
2. **Error Details**
100+
101+
![image.png](https://assets.emqx.com/images/b2ef52a41fa180dcfd67932836c505fb.png)
102+
103+
3. **Diagnostic Analysis**
104+
- **Error Description:** This error indicates that the system stopped execution as a safety measure due to an edit monitor timeout.
105+
- **Impact Scope:**
106+
- RAPID program execution was interrupted
107+
- Robot operation is paused
108+
- **Possible Causes:**
109+
- Watchdog commands were not sent periodically during RAPID and main program execution
110+
- RobotControlMate connection issues (if in use)
111+
- Misconfigured system monitoring parameters
112+
113+
4. **Recommended Solutions**
114+
115+
- **Immediate Actions:**
116+
**[Basic Fix]**
117+
118+
- Check and restart RAPID program execution
119+
- Verify RobotControlMate connection status (if applicable)
120+
121+
**[Configuration Adjustment]**
122+
*If not using RobotControlMate:*
123+
124+
1. Open the controller configuration interface
125+
2. Navigate to `Controller/OperatorSafety/Heartbeat`
126+
3. Set the parameter value to `"False"`
127+
4. Save changes and restart the system
128+
129+
- **Long-Term Prevention:**
130+
131+
- Regularly inspect the watchdog command mechanism
132+
- Establish a heartbeat monitoring log system
133+
- Consider upgrading to the latest control system software
134+
135+
- **Notes**
136+
- Back up current settings before modifying configuration parameters
137+
- Restarting may cause loss of unsaved program changes
138+
- Maintenance is recommended during non-production hours
139+
140+
Would you like more detailed guidance on RAPID program debugging or other related topics?
141+
142+
## **Related Technologies and Software**
143+
144+
The software used in this application is as follows:
145+
146+
![image.png](https://assets.emqx.com/images/7024315d9f20188aae8aa642901a2fa4.png)
147+
148+
- **Industrial Data Collection Software - NeuronEX:** Defines data collection and includes a northbound Sparkplug B plugin to report tag data to the cloud in Sparkplug B format.
149+
- **Message Broker - EMQX:**
150+
- Receives Sparkplug B data packets reported from the edge.
151+
- The Sparkplug App subscribes to relevant topics, storing the data in a time-series database.
152+
- **Time-Series Database - TDengine:** Stores tag data reported by devices and provides flexible data access capabilities.
153+
- **Relational Database - MariaDB:** Stores the mapping between devices and their nicknames, e.g., the device demo is nicknamed "Optimus Prime," facilitating the connection between OT and IT systems. In real-world business systems, this represents IT system data such as MES, ERP, and CRM.
154+
- **LLM - Deepseek v3:** Powered by Deepseek v3 from SiliconFlow.
155+
- **RAG Knowledge Base:**
156+
- Supports BAAI/bge-base-en-v1.5 and Aliyun vector models.
157+
- Uses a local MilvusVectorStore for the vector database.
158+
- User manuals are vectorized and stored in the vector database. When users query error codes, the application retrieves relevant content from the vector database, and the LLM generates solutions and recommendations based on this data.
159+
- **MCP:**
160+
- Encapsulates MCP services for TDengine time-series data to retrieve device-reported tag data, node status, and device online/offline information.
161+
- Encapsulates MCP services for MariaDB relational data to store mappings between OT device data and IT business data, e.g., the OT identifier demo corresponds to the IT nickname "Optimus Prime." Users typically query data using IT nicknames or aliases.
162+
163+
The relevant code has been open-sourced, and readers can access the source code at [GitHub - emqx/uns-demo: Use natural language to explore IIoT Sparkplug B data by leveraging MCP, AI and agent](https://github.com/emqx/uns-demo).
164+
165+
## **Summary**
166+
167+
This article briefly introduces how to leverage AI and various foundational software to build a cutting-edge IIoT data exploration application, providing insights for readers to develop their own AI-driven industrial applications. The EMQ industrial team is planning to officially integrate Sparkplug B data reporting, storage, and AI-based analysis into the EMQX ECP, with a release expected in the second half of the year. Interested readers are welcome to contact our sales team for early access.
168+
169+
170+
171+
<section class="promotion">
172+
<div>
173+
Talk to an Expert
174+
</div>
175+
<a href="https://www.emqx.com/en/contact?product=solutions" class="button is-gradient">Contact Us →</a>
176+
</section>
Lines changed: 45 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,45 @@
1+
为提升系统稳定性、性能和可观测性,我们将对 Serverless 平台进行升级。专有版(Dedicated)和 BYOC 部署不受影响。
2+
3+
## 升级时间安排
4+
5+
**升级时间:** 2025年6月3日 22:00-23:00(北京时间)
6+
7+
**预计中断:** 30-60秒(可能发生2-3次短暂断连)
8+
9+
**受影响服务:** Serverless 部署
10+
11+
## 升级影响
12+
13+
### 维护期间
14+
15+
- MQTT 连接将短暂中断 2-3 次,每次 30-60 秒
16+
- 暂停创建或修改 Serverless 部署
17+
- 无设备在线的工作负载不受影响
18+
19+
### 维护完成后
20+
21+
- 域名地址保持不变
22+
- IP 地址会发生变化
23+
- 如使用 IP 白名单,需及时更新
24+
25+
## 您需要做的准备
26+
27+
### 升级前
28+
29+
- 确保 MQTT 客户端支持自动重连
30+
- 建议使用域名连接,避免 IP 变更影响
31+
32+
### 升级后
33+
34+
- 如有 IP 白名单,请及时更新
35+
- 验证设备连接和数据传输正常
36+
37+
## 服务保障
38+
39+
- **实时状态:** https://status.emqxcloud.cn/
40+
- **技术支持:** 提交工单获得 SRE 团队支持
41+
- **严格流程:** 按标准维护流程执行,确保平稳过渡
42+
43+
我们将以最小的业务影响换取更好的服务体验。感谢您对 EMQX Cloud 的信任与支持!
44+
45+
**EMQX Cloud SRE 团队** **2025 年 6 月**

0 commit comments

Comments
 (0)