Digital Twin-Based Calibration System for Smoke Alarms

By Dr Silpaja Chandrasekar, PhDReviewed by Susha Cheriyedath, M.Sc.Nov 10 2023

In a paper published in the journal Scientific Reports, researchers addressed issues related to the outdated calibration method of smoke alarms, low production efficiency, and challenges in real-time monitoring. They proposed a method for modeling and monitoring smoke alarm calibration using a digital twin system. The approach began with analyzing the calibration requirements and outlined the framework for the digital twin calibration system for smoke alarms.

Subsequently, researchers developed a five-dimensional digital twin model of a smoke box, encompassing the physical smoke box, geometric model, physical model, rule model, and behavior model. The system's architecture, data acquisition, and mapping were employed to create the twin data model. Implementing the proposed method in an enterprise's calibration system validated its effectiveness, resulting in an improved success rate of smoke alarm calibration and a reduced rate of defective products.

Background

In fire safety and smoke alarm manufacturing, ensuring the sensitivity and quality of smoke alarms is paramount. Traditional manual calibration methods have limitations in terms of efficiency and operability. With the increasing demand for smoke alarms, there is a growing need for more efficient and data-driven calibration processes. Digital twin technology, which involves creating virtual models of physical entities, offers a promising solution to this challenge.

Overall Architecture Design

The digital twin cigarette box application framework comprises four essential layers: the physical layer, the twin layer, the data layer, and the service application layer. The physical layer consists of the material resources and resource perception. This layer includes components like the smoke box entity and smoke alarms at the physical resource layer. In contrast, the resource sensing layer involves elements like PLC, optical receivers, flow rate, and environmental sensors.

The physical smoke box entity collects real-time data from the physical smoke box via sensors and the TCP communication protocol, transmitting this data to the virtual smoke box system. The twin layer is at the system's core and encompasses the geometric, physical, and production behavior models. The data layer is responsible for data collection, which is then stored in the database, covering various data types. The service application layer integrates the twin and data layers, facilitating virtual simulation, real-time synchronization, and remote control.

Construction of Digital Twin Cigarette Box Model: Building the digital twin cigarette box model involves creating a five-dimensional digital twin that includes the physical cigarette box (PE), virtual cigarette box (VE), service (SS), twin data (DD), and connection (CN). Based on functionality and integration, the physical smoke box entity forms the foundation, divided into unit, system, and complex system levels. The virtual smoke box comprises geometric, physical, behavioral, and rule models. The geometric model reflects the physical entity's shape, size, position, and assembly relationships. The physical model incorporates attributes and properties of the physical smoke box. The behavior model includes sensor and actuator modeling to replicate real-world interactions. Finally, designers create the rule model to meet system control logic, ensuring the synchronization of the virtual twin with its physical counterpart.

Twin Data Model Construction: Data connects the physical and virtual worlds in the digital twin system. The communication system architecture involves multiple protocols, including OPC UA, to ensure seamless data flow. Data acquisition relies on physical sensors and Radio-Frequency Identification (RFID) readers to collect real-time information from the smoke box. This data is then processed and stored in a database. The central station communicates with the twin model, driving its operation by establishing a connection with virtual variables. Data mapping and driving encompass the smoke box, alarm, and environment mapping. This process enables real-time synchronization of equipment status data between the physical space and the digital twin, allowing them to operate in unison. Accurate data drives the virtual twin, ensuring dynamic consistency.

By implementing this comprehensive framework, the digital twin cigarette box system effectively replicates the physical smoke box's behavior, operation, and environmental conditions, leading to more efficient and accurate calibration processes.

Digital Twin-Based Smoke Alarm Calibration Validation

In this example, the calibration of a smoke alarm in a company located in Ningbo, Zhejiang Province, was used as a verification case to validate the modeling and mapping method presented in this paper. The modeling team used Siemens NX to model the calibration process of the smoke alarm, creating an active digital twin model of the smoke box.

They established the virtual-real communication platform using Object Linking and Embedding (OLE) for Process Control Unified Architecture (OPC UA) communication protocol. They employed real-time mapping technology to visualize the 3D model and virtual production.

The entire cigarette box, including its geometric relationships, assembly, and physical properties, was accurately modeled and compared to its physical counterpart. This digital twin allowed for real-time monitoring of the smoke alarm calibration process, enabling the collection and storage of production data in a database. Researchers compared and assessed the method's reliability between the traditional smoke alarm calibration method and the approach proposed in this paper. The conventional calibration method involved manual product placement, visual data observation, and lacked data collection during calibration.

Using the proposed method, the authors recalibrated 40 alarms that had previously undergone traditional calibration. The results showed that all products met the testing criteria of EN54-S/7:2000, demonstrating a significant increase in the qualification rate from 98% to 99.6%. Additionally, there was a notable 1.6% reduction in the repair rate for defective products, highlighting the effectiveness and improved performance of the designed system.

Conclusion

In summary, this paper explores digital twin technology in smoke alarm calibration. It presents the system's inner workings, proposes a comprehensive framework, and employs Siemens NX for a detailed five-dimensional model of the smoke box. Additionally, establishing a virtual-to-real communication platform through OPC UA and data collection techniques like RFID enhances efficiency and calibration quality. Ultimately, this research highlights how digitalization can transform smoke alarm production.