Unified Namespace Data Integrity

Solution Overview

The manufacturing industry is changing due to the integration of real-time data and centralized data management tools like the Unified Namespace (UNS) model, which stores different types of data in one location. Modern factories generate data from systems like enterprise resource planning (ERP), manufacturing execution system (MES), and shop floor machines. As manufacturers connect and automate their systems, unifying these data sources is crucial. For example, smart factory initiatives can boost productivity by up to 12 perfect and improve equipment effectiveness by up to 20 percent, according to Deloitte.

This solution implements a demo named Leafy Factory that develops a comprehensive UNS framework with MongoDB. You can use MongoDB as a UNS that autonomously unifies and analyzes data from different manufacturing systems. MongoDB's flexibility, real-time processing, and scalability make it an optimal choice for an efficient UNS. It also brings improved efficiency and insights to manufacturing data management systems, as shown in the following diagram:

click to enlarge

Figure 1. Automation pyramid versus a Unified Namespace

This UNS framework ingests diverse operational data types, analyzes streams for actionable insights, stores information in MongoDB, and provides strategic recommendations using comprehensive analytics generated from different data sources. This allows for increased efficiency by centralizing data storage and management.

Reference Architectures

The following diagram displays this solution's architecture:

Figure 2. Leafy Factory UNS architecture

This architecture has the following data flow:

1. Data Ingestion with MQTT Broker

The solution begins by employing Cedalo's Mosquitto MQTT Broker to handle real-time data streams from shop floor machines. This broker gathers data such as machine status and sensor readings, including temperature and vibration. The architecture remains broker-agnostic and is capable of integrating with various other MQTT providers as needed.

2. SQL Data Integration via Kafka Connector

Concurrent with the MQTT data reception, the Debezium connector captures SQL data from ERP and MES systems, such as work orders and material tracking, and streams it into Kafka topics. Kafka connector then processes this data and inserts it into MongoDB Atlas. This ensures that ERP data is continuously updated.

3. Database Management with MongoDB Atlas

MongoDB's flexible document model allows you to store diverse data structures, from raw machine sensor data to structured ERP records, in Atlas. By storing data in a document format, manufacturers can effortlessly adapt to changes, such as new sensors or machine attributes.

The data in this solution has the following structure:

{
   "result": {
      "factory": {
         "location": "qro_fact_1",
         "timestamp": "2025-04-12 02:59:41.569745",
         "production_lines": [
         {
            "production_line_id": 2,
            "machines": [
               {
               "_id": 3,
               "machine_id": 3,
               "details": {
                  "machine_status": "Available",
                  "last_maintenance": "2024-10-31 14:25:00",
                  "operator": "Grace Conway",
                  "avg_temperature": 70.48,
                  "avg_vibration": 0.59,
                  "temp_values": 70,
                  "vib_values": 0.01
               },
               "work_orders": [
                  {
                     "id_work": 111,
                     "jobs": [
                     {
                        "id_job": 62
                     }
                     ]
                  },
                  {
                     "id_work": 105,
                     "jobs": [
                     {
                        "id_job": 58
                     }
                     ]
                  },
                  {
                     "id_work": 104,
                     "jobs": [
                     {
                        "id_job": 57
                     }
                     ]
                  },
                  {
                     "id_work": 100,
                     "jobs": [
                     {
                        "id_job": 55
                     }
                     ]
                  },
                  {
                     "id_work": 99,
                     "jobs": [
                     {
                        "id_job": 52
                     }
                     ]
                  }
               ]
               }
            ]
         }
         ]
      }
   }
}

4. Real-time Analysis with Time Series Collections

Once the data is in Atlas, you can use time series collections to manage and analyze your data. Time series collections allow manufacturers to store time-stamped data and gain insights into manufacturing processes over time, allowing you to spot patterns and react promptly to any changes. You can also integrate additional metadata into the document model, such as machine ID, operator name, location, and more.

5. Visualization and Analytics with Atlas Charts

Finally, you can visualize the integrated data by using Atlas Charts, which provides intuitive graphical representations of production metrics, quality analysis, and machine statuses. This enables manufacturers to make informed decisions through actionable insights.

Build the Solution

This solution uses a combination of core technologies that work together to allow seamless data processing and integration across diverse manufacturing systems. For complete implementation details, including code samples, configuration files, and tutorial videos, see this solution's GitHub repository.

This solution has the following prerequisites:

• Python 3.12 or later

• Node.js 14 or later

• MongoDB Atlas Cluster running MongoDB 8.0.4 or later

• Apache Kafka 3.9.0 or later

• Java JDK 23 or later

• PostgreSQL 15.10 or later

To deploy the solution, follow the instructions in the README in the GitHub repository linked above.

Key Learnings

• Adaptability to operational changes: MongoDB's flexible document model can easily integrate new data sources and scale with expanding production lines, allowing manufacturers to maintain a dynamic data system without requiring architectural overhauls.

• Centralized data layer for advanced applications: While the UNS doesn't directly perform applications like predictive maintenance, it provides infrastructure for such initiatives. Manufacturers can easily implement IoT-based solutions, enhance maintenance schedules, and optimize costs by using the UNS to centralize real-time and historical data.

• Cross-functional insight generation: MongoDB's analytics capabilities allows manufacturing teams to integrate diverse datasets, such as MES metrics and ERP outputs.

• Ensured data availability and reliability: MongoDB's replica set architecture ensures that the system remains operational without interruption. This prevents potential disruptions in the manufacturing data ecosystem and supports reliable long-term operations.

Authors

• Raphael Schor, MongoDB

• Romina Carranza, MongoDB

• Giovanni Rodriguez, MongoDB

Learn More

• Build an IoT Data Hub for Smart Manufacturing

• Framework for Rapid AI Agent Deployment

• Elevate Flight Operations with Real-Time Analytics