Abstract
Modern manufacturing supply chains operate in an increasingly volatile environment characterised by geopolitical uncertainty, fluctuating demand, supplier instability, transportation disruptions, resource constraints, and rapidly changing market conditions. For furniture manufacturers, these challenges are amplified by complex multi-tier supply networks, long material lead times, dependency on external suppliers, and the continued use of manual planning and monitoring processes. Traditional supply-chain management approaches are often reactive, identifying disruptions only after they have already negatively affected production or delivery performance.
This article presents NARRATE’s Real-Time Risk Identification & Monitoring System (RIMS) designed to support proactive risk management within smart manufacturing environments. The system combines heterogeneous internal and external data and event sources provided by the consortium such as real-time messaging, Complex Event Processing (CEP), analytics, enterprise modelling, and dashboard-based decision support. In addition, a Predictive Market Analysis and Early Warning Tool continuously analyses global news data to identify economic, geopolitical, and trade-related events that may affect industrial value chains.
RIMS extends the capabilities of the Intelligent Manufacturing Custodian (IMC) by transforming diverse operational and external information into structured risk intelligence and actionable alerts. The resulting architecture enables manufacturing organisations to move from retrospective monitoring towards continuous risk identification, early warning, and proactive decision-making. The approach is particularly relevant to furniture manufacturing, where improved supplier visibility, inventory management, logistics coordination, and disruption preparedness can directly contribute to operational resilience, sustainability, and customer satisfaction.
Keywords: Smart Manufacturing, Supply-Chain Resilience, Risk Identification, Early Warning Systems, Complex Event Processing, Predictive Analytics, Enterprise Modelling, Industrial IoT, Furniture Manufacturing
Introduction
Supply-chain resilience has become a critical requirement for modern manufacturing organisations. Globalised production networks provide access to specialised suppliers, materials, logistics services, and international markets, but they also increase exposure to external disruptions. Geopolitical conflicts, trade restrictions, energy-price fluctuations, transportation bottlenecks, natural disasters, supplier failures, and unexpected changes in demand can propagate rapidly across interconnected supply networks.
The furniture industry is particularly exposed to these challenges because manufacturing depends on a broad range of materials, components, suppliers, logistics providers, and production resources. Disruptions affecting a single material or supplier can consequently result in production delays, increased inventory requirements, missed delivery commitments, or lost sales. These challenges are further intensified in companies where supply-chain planning and monitoring remain dependent on manual processes. Enterprise Resource Planning (ERP) systems may contain extensive operational information, but this information is frequently distributed across different modules and systems and may not be transformed into real-time risk intelligence. Similarly, external events that could affect supply chains are typically monitored manually or considered only after their consequences become visible.
The solution presented in this article addresses this gap through a Real-Time Risk Identification & Monitoring System (RIMS). The objective is not simply to visualise existing operational information but to continuously combine, process, contextualise, and interpret heterogeneous signals which aim to identify emerging risks at an early stage. The system forms part of a broader Smart Manufacturing as a Service (MaaS) approach and extends the functionality of the Intelligent Manufacturing Custodian (IMC). It provides a technical foundation for connecting operational manufacturing information with external market intelligence and converting these inputs into actionable decision-support information.
Problems of Vulnerable Manufacturing
Resilient manufacturing organisation must be able to continuously overview their manufacturing and supply-chain environments while keeping eyes on possibly emerging, disruptive events and corresponding mitigation actions. A resilient manufacturing organisation must be able to address these three fundamental aspects to enable informed, proactive decisions that build resilience and protect value (c.f. figure below).
Conventional monitoring solutions primarily address the first challenge. They provide information about inventory, production status, equipment condition, supplier performance, and logistics. However, they often provide only limited support for identifying relationships between these indicators and external events.
For example, an increase in geopolitical tensions may initially have no visible effect on a furniture manufacturer. However, if the event subsequently affects trade routes, material availability, commodity prices, or a key supplier, operational consequences may appear several weeks later. A purely operational monitoring system may hence only detect the resulting shortage once inventory levels begin to fall.
RIMS introduces an additional layer of intelligence by connecting operational indicators with external disruption signals. This allows the system to identify potential risks before they become operational incidents and addresses hereby the principal technical challenges as pinpointed in the figure below.
System Concept
RIMS is designed as a modular architecture in which data acquisition, event processing, analytics, artificial intelligence, enterprise modelling, visualisation, and alert dissemination operate as interconnected components. At a conceptual level, the system integrates the pipeline as sketched below.
The orchestration of the above-mentioned system pipeline provides the following functionalities:
- Data Acquisition: collects information from internal and external sources. The heterogeneous nature of these sources requires an architecture capable of handling both structured and unstructured data. Structured operational information can be obtained from ERP systems, databases, equipment monitoring systems, and IoT platforms. External information is obtained from large-scale news databases.
- Data integration: Responsible for contextualizing these data streams and integrate available data to downstream processing components while preserving the temporal characteristics of real-time information.
- Event processing: Continuous real-time streams of industrial information contain a large number of events, many of which are operationally insignificant. A Complex Event Processing (CEP) engine provides mechanisms for detecting meaningful combinations and temporal patterns within these streams.
- Risk identification is also handled via the CEP to identify conditions such as abnormal equipment behaviour, repeated supplier-performance degradation, inventory falling below defined thresholds, increasing delivery delays, etc. The advantage of CEP is that risk identification can occur continuously rather than through periodic manual batch analysis.
- Contextual analysis: Enterprise Modelling provides an additional architectural foundation for RIMS. Manufacturing environments consist of interconnected processes, resources, information flows, organisational units, suppliers, and technologies. Representing these relationships explicitly enables risk information to be interpreted within its operational context.
- Alert generation is obtained on one side via the event processing and event correlation. In addition, the Predictive Market Analysis and Early Warning Tool (GDELT) identifies relevant news, processes it using the analytical pipeline, and generates additional structured alerts.
- Decision support: NARRATE’s Intelligent Manufacturing Custodian (IMC) and monitoring dashboard present the information to relevant users.
The underlying system, architecture combines two complementary information domains – the internal operational information basis and external sources. Internal operational information describes the current state of the manufacturing and supply-chain environment (c.f. figure below).
These sources provide the operational context required to determine whether an external or internal event represents a meaningful risk. External intelligence on the other hand provides information about events outside the organisation that may affect its operating environment.
The integration of these two domains is a central feature of the proposed approach.
Application to Furniture Manufacturing
The relevance of the architecture is illustrated through a furniture-manufacturing scenario. Consider a manufacturer depending on several suppliers for materials and components. Under normal conditions, the ERP system provides information about inventory, purchase orders, supplier performance, and production requirements. A conventional system may detect a problem only when inventory falls below a critical level or when a supplier misses a delivery.
With RIMS, the monitoring process can begin earlier: Suppose global news indicates a developing geopolitical or trade event that could affect the availability or transportation cost of a particular material. The Predictive Market Analysis and Early Warning Tool identifies relevant news, processes it using the analytical pipeline, and generates a structured alert. The system can then combine this external signal with internal information such as:
- current material inventory;
- supplier lead time;
- historical supplier reliability;
- open purchase orders;
- production requirements; and
- affected products.
The resulting risk assessment can therefore be considerably more informative than either the external news signal or the ERP information considered independently. This enables managers to investigate potential alternatives before the disruption reaches production.
Benefits & Conclusion
The Real-Time Risk Identification & Monitoring System provides a technical foundation for proactive risk management in smart manufacturing environments. By combining heterogeneous operational data, real-time messaging, complex event processing, enterprise modelling, and external market intelligence, the system transforms large and diverse information streams into structured, actionable risk intelligence.
The integration of RIMS into a smart manufacturing environment provides several potential benefits.
- Operational resilience: Early identification of supply-chain risks allows organisations to prepare mitigation measures before disruptions affect production.
- Supply-chain visibility: Combining internal and external information provides a broader view of the factors influencing supply-chain performance.
- Reduced reaction time: Automated monitoring and alert generation reduce the dependency on manual information collection and analysis.
- Improved supplier management: Supplier performance can be evaluated alongside external events, helping organisations distinguish isolated supplier problems from broader market or geopolitical risks.
- Inventory optimisation: Earlier risk identification can support more informed inventory decisions and reduce the need for excessive safety stock.
- Improved delivery performance: Potential disruptions can be identified before they translate into missed production schedules or customer deliveries.
- Sustainability: Better planning and reduced disruption can contribute to more efficient resource utilisation, reduced emergency logistics, and lower levels of unnecessary inventory and waste.
- Enhanced managerial decision-making: The combination of structured operational information, external intelligence, enterprise modelling, and AI-supported analysis gives managers a more comprehensive basis for intervention.
The Predictive Market Analysis and Early Warning Tool significantly extends capability of conventional systems by introducing continuous monitoring of global economic, geopolitical, and trade-related developments. Rather than waiting for external disruptions to become visible through declining inventory, supplier delays, or production problems, the system provides an opportunity to identify relevant signals at an earlier stage. Integration with the Intelligent Manufacturing Custodian further connects risk identification with manufacturing decision support. This creates a pathway from external event detection to contextualised operational assessment and, ultimately, preventive action.
The proposed architecture therefore represents a transition from reactive supply-chain monitoring to proactive, intelligence-driven resilience management. By connecting real-time manufacturing information with external market intelligence and enterprise context, RIMS establishes an important technological foundation for resilient, sustainable, and increasingly autonomous manufacturing operations.
Authors
