Integrating an ERP into Unified Namespace

ERP systems, traditionally the backbone of organizational data management, are designed to streamline and automate core business activities. However, the dynamic and interconnected nature of modern manufacturing demands more than what standalone ERP systems can offer. 

The Unified Namespace has proven to be the architectural approach that meets the demands for a more connected, intelligent, and adaptable digital ecosystem, driving innovation and competitive advantage in digital manufacturing. This article explores how to integrate the ERP system into a Unified Namespace data ecosystem. 

Why Integrate ERP Data into Unified Namespace

Manufacturing companies often store their organizational master data model in an ERP system. However, this data typically doesn't accurately reflect the company's entire digital infrastructure, making it unreliable as a single source of truth — among other reasons.

Traditionally, to integrate ERP data — like product codes, work orders, and schedules — with a Manufacturing Execution System (MES), companies use several methods. One method involves the MES making REST API calls to retrieve data from the ERP, requiring familiarity with the ERP's REST API documentation. 

Another method is connecting through an SQL database, where the MES would use an SQL connector, such as ODBC, necessitating knowledge of the ERP's database structure. Lastly, a custom connector could be used, acting as an MES extension to link with the ERP. These integration methods are difficult to maintain, especially when system modifications are made, as they require manual updates to the data integration processes.

The Unified Namespace (UNS) architecture offers a more streamlined solution. It uses the organizational structure from the ERP to create a semantic hierarchy. This hierarchy is then published to an MQTT broker, allowing ERP information to be integrated into the UNS at appropriate levels, along with data from other systems across the industrial technology stack.

In this architecture, ERP and MES systems are allocated their namespaces within the semantic hierarchy, facilitating data exchange through a unified interface without needing to understand each other's implementation specifics. Changes made in either the ERP or MES, such as adding or removing functions, are automatically updated in the namespace.

Moreover, this approach ensures that the data is accessible for future use by other systems, enhancing efficiency and adaptability across the organization's digital infrastructure.

Connecting ERP Systems to a Unified Namespace

Most ERP systems available today do not inherently support the MQTT communication protocol. As a result, integrating these ERP systems with a Unified Namespace typically requires an additional layer, the DataOps layer. This layer serves as an intermediary, facilitating the conversion of ERP-specific data structures into MQTT-compatible formats and performing necessary data processing. This ensures the data is correctly formatted and placed within the Unified Namespace.

The process involves utilizing an Industrial Internet of Things (IIoT) platform that acts as an informational bridge. This platform connects to the ERP system through a REST API endpoint or SQL database connector, converting and relaying data to the Unified Namespace via MQTT.

Once this infrastructure is established, Manufacturing Execution Systems (MES) can subscribe to the ERP's namespace to pull needed data directly from the ERP system. Conversely, the ERP system can subscribe to the MES namespace to receive information sent from the MES. This two-way communication streamlines data exchange, allowing both systems to stay synchronized without the need for direct knowledge of each other's internal data structures.

Publishing ERP Data to Unified Namespace

Publishing ERP (Enterprise Resource Planning) data into a Unified Namespace (UNS) requires a method that is both adaptable and organized, designed to meet the unique requirements of the business and its operational processes. Using the ISA-95 data model hierarchy — comprising Enterprise, Site, Area, Line, and Cell — to structure the UNS MQTT topic hierarchy, ERP data can be integrated at various levels. It's common to assign ERP namespaces at the plant level, or within specific areas of larger plants, allowing ERP systems to disseminate data, often extending to the Line level.

These ERP Namespaces are repositories for data like Production Orders, Bill of Materials (BOM), and Inventory. The details of ERP data shared within the UNS can significantly differ across production settings. In some scenarios, Work Orders might be issued to a particular ERP topic as a dataset, then segmented into individual tags.

An alternative method involves treating each Work Order as a separate Topic within a clearly defined path, such as Enterprise/Site/ERP/Orders, with distinct topics for the BOM and material IDs. This setup allows for periodic updates, replacing the set of jobs as necessary. 

This flexible approach also applies to managing job statuses. Starting a job leads to it being published in an appropriate Topic path like Enterprise/Site/Area/Line/CurrentJob. This strategy highlights the necessity of customizing the UNS framework to align with the specific operational needs of the business.

Example ERP Interaction Through Unified Namespace

To illustrate the dynamic interaction between ERP integration and a Unified Namespace (UNS) in a manufacturing context, let's consider the following example, which outlines a sequence of events from the creation of an order to the completion of a product:

New Flavor Launch

The ERP system records the introduction of a new flavor — 'Citrus Zing'. The details of this new product, including the recipe, batch size, and packaging requirements, are published to the UNS under a specific product ID. This ensures that all relevant systems are aware of the new production order.

Production Scheduling

A scheduling tool picks up the new flavor order published to the ERP namespace within the UNS. It evaluates production line availability, ingredient stock levels, and existing orders to generate an optimized production schedule. This schedule is then published back to the UNS for access by the relevant departments.

Ingredient Preparation

The Warehouse Management System (WMS), upon receiving the updated schedule through the UNS, identifies and arranges for the delivery of necessary ingredients and packaging materials to the designated production line well ahead of the start time.

Production Assignment

The Manufacturing Execution System (MES), now updated with the latest schedule, assigns the 'Citrus Zing' production order to a specific bottling line, tagging it under 'Available Orders' for that line's namespace within the UNS.

Kickstarting Production

Operators at the bottling line activate the 'Citrus Zing' order via the MES interface. This triggers updates in the UNS, instructing the PLCs on the bottling line to adjust settings for the new flavor profile.

Ingredient Tracking

As the production starts, each ingredient batch used is scanned and verified at the production line. These transactions are automatically recorded in the ERP through the UNS, providing real-time tracking of ingredient usage and inventory levels.

Production Monitoring

Throughout the bottling process, PLCs monitor various parameters such as fill levels, cap integrity, and labeling accuracy. This data is continuously published to the UNS, allowing for real-time monitoring and adjustments as needed to maintain quality standards.

Conclusion

In conclusion, the integration of ERP systems into a Unified Namespace represents a significant advancement in the digital transformation journey of manufacturing companies. By bridging the gap between traditional ERP data management and modern, interconnected digital ecosystems, organizations can achieve unprecedented levels of efficiency, agility, and insight.