WHITE PAPER
MQTT Platform for AI: Empowering AI with Real-Time Data →

Overview

A renowned Fortune 500 food manufacturing company, known for its popular chocolates, biscuits, and beverages, operates in over 150 countries. One of its facilities, equipped with 15 automated production lines, embraces cutting-edge technologies like AI, IoT, big data analytics, and digital twins to optimize operations and lead the industry in smart manufacturing.

In a facility as advanced as this, operational continuity relies on the seamless functioning of thousands of rotating and moving components. Any production outage can halt production, leading to significant financial losses and potential safety risks. To stay ahead of potential equipment failures, the company sought to implement a predictive maintenance platform—one that could foresee issues before they occurred, ensuring smooth production and optimized equipment performance.

The Challenge: Beyond Automation, Toward Intelligence

With thousands of machines producing high-frequency data every second, the key challenge was not just collecting data but making sense of it in real time. The company needed a platform capable of:

  • Seamless Device Integration: Their equipment, ranging from sensors and PLCs to CNC machines, used diverse communication protocols, creating a fragmented data landscape.
  • Real-time Data Processing: Predictive maintenance required ultra-high-frequency data collection—multiple times per second from each device—while ensuring this data could be processed and acted upon in real time.
  • Effective Data Management: Handling such vast volumes of time-series data with precision, storing it securely, and making it accessible for analysis presented significant technical hurdles.
  • Unified Operations: The solution needed to harmonize with existing smart manufacturing systems, providing engineers with an intuitive interface to monitor equipment, predict failures, and take preventive actions.

The Solution: Building a Predictive Maintenance Platform with EMQ

To meet these challenges, the company partnered with EMQ and WYSEngine to build a scalable, cloud-edge predictive maintenance platform. EMQ's software-defined IoT solution, combined with WYSEngine's advanced AI algorithm, provided the necessary capabilities for data collection and processing to effectively manage the factory's complex data environment and gain essential insights.

Seamless Integration of Diverse Devices with High-frequency Data Collection

EMQ’s NeuronEX industrial gateway supports over 80 industrial protocols, enabling seamless connectivity for thousands of devices. Whether it is temperature, pressure, or vibration data, NeuronEX ensures that all sensor data is unified and transmitted via the lightweight MQTT protocol, maintaining real-time visibility even in low-network conditions. It also offers 100ms-level high-frequency data collection. Through its visual configuration tools, NeuronEX enables unified data naming and standardization, ensuring that the predictive maintenance platform receives high-quality, structured data essential for accurate and timely fault detection.

Architecture diagram

Real-time Edge Computing for Instant Insights

NeuronEX also provides edge computing capabilities, allowing critical data processing to occur near the devices themselves. This enables low-latency anomaly detection and real-time analysis, reducing the load on the central system and boosting the accuracy and speed of predictive algorithms.

High-Performance Data Streaming for Predictive Maintenance

EMQX, EMQ’s enterprise-grade MQTT platform, ensures reliable, low-latency data streaming. Capable of handling over 100,000 transactions per second, it enables seamless real-time transmission of high-frequency equipment data into time-series or relational databases. This capability supports the creation of high-precision historical data assets for predictive maintenance, providing a robust foundation for complex predictive models and fault detection algorithms.

The Impact: Transforming Maintenance, Boosting Efficiency

Within just two weeks, the Suzhou facility’s predictive maintenance platform was fully operational, significantly faster than anticipated—improving project efficiency by 35%. Powered by EMQ, the platform now monitors thousands of critical data points, with data refresh rates as fast as 100 milliseconds. This real-time visibility into machine performance enabled the company to detect potential issues long before they could impact production.

Since its deployment, the platform has delivered remarkable results:

  • 83% reduction in unexpected downtime: By predicting equipment failures before they occurred, the company dramatically reduced unplanned shutdowns, avoiding costly interruptions to production.
  • 8.7% improvement in production efficiency: With equipment running more smoothly and maintenance being performed proactively, the entire production process became more streamlined and efficient.
  • Future-proofing with flexibility: Beyond predictive maintenance, the EMQ solution serves as a foundation for the company’s broader digital transformation strategy. It can easily integrate with other systems, such as MES and ERP, paving the way for a holistic, connected manufacturing ecosystem. This adaptability has shortened implementation times for new projects by 30% and reduced costs by over 20%.

A New Era of Intelligent Manufacturing

With EMQ’s cutting-edge industrial IoT platform, this global food manufacturing leader has not only enhanced its equipment management but also set the stage for long-term operational excellence. The predictive maintenance platform has empowered the company to move from reactive to proactive operations, driving both productivity gains and cost savings. As they continue to expand their digital footprint, the partnership with EMQ remains a critical pillar in their journey toward intelligent, data-driven manufacturing.