Building a Smart Water Plant with EMQX MQTT Platform and Digital Twin Technology

Introduction

Utilizing digital twin technology in conjunction with IoT, artificial intelligence, and big data, a digital twin platform for water management can help to gain a deeper understanding of the operational status of water systems, enhancing system security, stability, and efficiency while optimizing water resource utilization and operational costs. Additionally, it provides decision-makers with a more scientific basis and decision support, further advancing the digitization and intelligence of the water industry.

Leveraging the lightweight MQTT messaging protocol and a cloud-edge collaborative data processing architecture, EMQX MQTT Platform delivers a data ingestion capability of 100,000 messages per second and data transmission with latency within 100ms. This results in an over 40% improvement in the real-time collection and processing efficiency of water management data.

Challenges of Building a Smart Water Plant

Heterogeneous Communication Protocols and Complex Network Environments

A water plant involves various industrial devices and systems, each with its own communication protocol or format. Additionally, weak network conditions may occur due to the distance of production equipment. Establishing a high-concurrency, low-latency communication channel in such a complex environment is crucial for real-time data perception.

Streaming Data Processing

In order to overcome the limitations in network, storage, and computational resources, a digital twin platform should provide capabilities for data cleansing, preprocessing, and real-time logical processing at the edge. This allows the digital twin to be presented based on real-time perception data and device status during the transmission of massive amounts of data.

High-Precision Data Storage

The core of digital twins lies in algorithms, and the effectiveness of their implementation depends primarily on the scientific nature of the algorithms and the precision of historical data. Faced with massive, high-frequency, and structurally complex water management data, persistently storing this data in databases in real-time and efficiently, whether aggregated in private or public clouds, poses a significant challenge.

Building a Digital Twin Smart Water Plant with EMQX MQTT Platform

EMQX MQTT platform, coupled with the NeuronEX gateway, enables the construction of a digital twin smart water plant. This system facilitates comprehensive monitoring, timely data processing, efficient storage, and analysis, ultimately leading to enhanced operational efficiency, reduced maintenance costs, and the assurance of water quality in the water plant.

Device Data Collection

The NeuronEX industrial gateway can connect various devices, sensors, and monitoring tools in the water plant to the system near the device end, transmitting data through the MQTT protocol. EMQX's high-availability cluster can handle large-scale device connections and data transfers, ensuring real-time and reliable data.

Data Processing at the Edge

NeuronEX can use edge computing for real-time data processing and analysis near the device end. This reduces latency and enables quick detection and prediction of anomalies.

Data Storage and Processing

EMQX provides support for the efficient collection and transmission of device data to cloud or local servers. It can handle a message rate exceeding 100,000 per second, making it ideal for managing and analyzing water quality indicators, flow, pressure, and other related data in real time.

Digital Twin Model Establishment

Leveraging millisecond-level real-time data collected by EMQX and standard MQTT, HTTP data interfaces, digital twin models of the water plant can be easily established using 3D visualization technologies such as BIM and Unity 3D. These models provide a real-time reflection of the water plant's operational status.

Operational Prediction and Optimization

EMQX provides high-precision historical data assets that can be combined with machine learning, deep data mining, and other AI algorithms to optimize and predict the operation of water plants. For instance, by analyzing models, it is possible to determine the most efficient water treatment processes, equipment operation schemes that save energy, and predict equipment failures or water quality anomalies based on real-time monitoring data.

The Advantages of EMQX MQTT Platform

All-in-One: Data Collection, Processing, Transmission, and Storage

EMQX, combined with NeuronEX, can realize the collection of water plant equipment data, edge computing, real-time data transmission, and data persistence. Through a decoupled approach, it provides the foundation for high-precision simulation, business integration, prediction and optimization for the digital twin platform.

Cloud-Edge Stream Data Processing

Both the NeuronEX at the edge and the EMQX in the cloud have the ability to logically process data through a rule engine. They can independently or collaboratively achieve real-time computation analysis, standardize messages, filter and clean data, intelligent alarms, and business routing at different levels. This significantly enhances the system's data processing and abnormal event perception capabilities.

Powerful Scalability

EMQX has open interfaces, horizontal scalability, and configurable modules that support flexible device integration. This makes it easier to respond to new business data needs, and helps enterprises expand rapidly and cost-effectively.

Related Products

EMQX Enterprise

EMQX Enterprise is a powerful enterprise MQTT platform designed for large-scale deployments and high reliability in IoT applications. The following capabilities of EMQX Enterprise can benefit the industry:

• High Reliability and Scalability: EMQX Enterprise adopts a distributed architecture with high availability and scalability to handle large-scale concurrent message transmission. It supports horizontal scaling to accommodate the growing number of IoT devices and data traffic, ensuring system stability.
• Rich Protocol Support: EMQX Enterprise supports multiple messaging protocols besides the MQTT protocol. It allows developers to extend to support all kinds of private protocols for their application needs.
• Data Integration: EMQX Enterprise seamlessly integrates with various data storage services, message queues, cloud platforms, and applications. It can connect with cloud services, enabling remote data transmission and cloud-based analysis.
• Security and Authentication: EMQX Enterprise provides robust security features, including TLS/SSL encrypted transmission, client authentication, and access control. It supports various authentication methods such as username/password, X.509 certificates, and OAuth, ensuring the security of IoT communication.
• Rule Engine and Data Processing: EMQX Enterprise has a flexible rule engine for real-time data processing and forwarding based on device data. It supports operations such as data filtering, transformation, aggregation, and persistence, helping users analyze and make decisions based on their business needs.
• Visual Monitoring and Management: EMQX Enterprise provides an intuitive visual monitoring and management interface, allowing users to monitor IoT devices and message transmission in real-time. Users can view connection status, message traffic, and other metrics and perform device management, troubleshooting, and system configuration operations.

NeuronEX

NeuronEX is software tailored for the industrial sector, focusing on equipment data collection and edge intelligent analysis. It is primarily deployed in industrial settings, facilitating industrial equipment communication, industrial bus protocol acquisition, industrial system data integration, edge-level data filtering and analysis, AI algorithm integration, and integration with IIoT platforms. It offers the following features:

• Diverse Connectivity: NeuronEX provides multi-protocol access capability, supporting simultaneous access to dozens of industrial protocols such as Modbus, OPC UA, Ethernet/IP, BACnet, Siemens, Mitsubishi, and more. It enables integration with multiple data sources from systems like MES and WMS within the enterprise. NeuronEX also supports bidirectional data flow, allowing both data acquisition and control command issuance to devices.
• Deploy Anywhere: NeuronEX has very low memory footprints and is suitable for running on low-profile architecture devices like x86, ARM, RISC-V, etc. It also supports docker-like containerized deployment, running with other co-located containers in Kubernetes environments.
• Data Processing: NeuronEX incorporates 100+ built-in functions, supporting data filtering, data manipulation, device control, and data persistence, storing data in a time-series database.
• Algorithm Integration: NeuronEX supports the integration of algorithms written in languages such as C, Python, and Go. It facilitates real-time inference of industrial mechanism models, machine learning, and deep learning models at the edge, enabling alerts and intelligent decision-making.
• Edge to Cloud: Through protocols like MQTT and SparkplugB, NeuronEX aggregates and pushes industrial data to cloud platforms. The bidirectional data flow between NeuronEX and the cloud platform establishes cloud-edge data coordination and control synergy, leveraging the platform's big data storage and analysis capabilities to amplify the value of NeuronEX usage.