Real-time data monitoring
Real-time data monitoring component provides a system for receiving events produced by sensors. This S2CP component can collect information generated in a city, farm or other food-related organization, so that information can be later represented for real-time visualization or monetization.
Functional architecture
The functional architecture shows that different types of sensors can supply data to real-time data monitoring, such as WiFi/Ethernet sensors, and other sensors that communicate with well-stablished technologies such as LoRA or SigFox. A connector-based adaptation middleware serves as a communication proxy and transmits information to two types of brokers:
- The semantic broker BSE publishes the interfaces for data consumption through a semantic model compatible with the data integration and management tool, and based on semantic models from previous projects such as DEMETER.
- The Orion Context Broker is a FI-WARE component optimized for sensor data communication, which supports the NGSIv2 protocol.
To provide persistence to the information received by the sensor infrastructure, the QuantumLeap module is used, another FI-WARE component that saves all sensor updates in a CrateDB sensing database.
Finally, this database can be used as a source of information for the representation of the information received in a dashboard, for subsequent analysis, filtering and as a decision support system.
Conducting the experiment
The hypothesis to validate is that by using the S2CP platform, it is possible to reduce the number of thermal discomfort episodes of food processing personnel by 10%, and it is possible to generate alerts related to realtime information in less than 30 seconds since error detection.
Variables such as temperature, humidity, power consumption and air quality were monitored through sensor nodes. A framework of performance and economic indicators is generated, and we analyze their changes and envisioned evolution after the Industry 4.0 paradigm adoption.
To achieve that, the real-time data monitoring component developed by UPM-P20 was used, as it can receive events produced by sensors and represent the mesaures for real-time visualization or monetization. As stated in Section 8.2 this component requires some data input coming from an Internet of Things (IoT) hardware platform, which must be defined and deployed in the pilot site. The output of this component will be a database containing all captured real-time information, expressed in a semantic and interoperable format (according to data models defined in T6.2), and also a visualization dashboard to facilitate data visualization, filtering and some simple analysis.
The following figure shows the deployed platform:
All software components were deployed in a private cloud in the Universidad Politécnica de Madrid. The underlying server was a Linux architecture (Ubuntu 20.04 LTS) with the following hardware characteristics: Dell R540 Rack 2U, 96 GB RAM, two processors Intel Xeon Silver 4114 2.2GHz, HD 2TB SATA 7,2K rpm.
