WISE: Low-Power WIrelesS IoT platform with Edge computing and predictive analysis for steam traps monitoring

This project presented a novel Internet of Things (IoT) device adapted for Industry 4.0 applications. The key innovation of this device is its battery-free and heat-powered design, addressing the restrictions associated with the use of lithium batteries in high-temperature and explosive industrial environments. Therefore, this sensor platform is a promising candidate for effective process monitoring in these demanding conditions such as the example case presented by the steam traps.

The advantages of the device over existing technologies have been demonstrated, including autonomy, maintenance-free operation, edge computing capabilities, high data transmission rates, minimal intrusiveness and easy installation.

The solution presented solution that uses waste heat to power IIoT devices in complex industrial environments, such as oil refineries and chemical plants. The proposed system uses thermoelectric technology to eliminate the need for batteries, significantly reducing maintenance costs and ensuring compatibility with explosive atmospheres (ATEX). A key advantage of the NB-IoT protocol is its ability to provide long-range, low-power communication, which is especially advantageous in large facilities such as refineries, where extensive wiring is impractical. The protocol ensures efficient data transmission with low latency, scalability and high reliability, enabling effective real-time monitoring in large-scale industrial operations. In addition, an in-depth study of the power production of the thermoelectric generator was conducted, which showed that the energy generated is well adapted to meet the energy demands of the IIoT node. This ensures consistent and sustainable operation of the system without relying on external energy sources. In addition, the study compared different data transmission methods, including MQTT + TLS and UDP, assessing their impact on energy consumption. UDP was found to offer significant power savings over MQTT + TLS, making it a more efficient choice for low-power, long-range communications in IIoT applications. Although the primary focus of this study is the remote transmission of data for analytics in the cloud using NB-IoT, we recognize that most critical industrial environments rely on direct SCADA integration due to security concerns under the ISA95 hierarchy. Future developments could explore multiprotocol support to facilitate both cloud and on-premises data management.

In the context of steam leak detection in traps, a major challenge lies in selecting the right leak detection method. In some cases, relying solely on temperature measurements is insufficient, especially since temperature data may not always be available in on-site operations. Therefore, additional parameters, such as ultrasonic measurements, are essential to improve the accuracy of the system. Ultrasound technology makes it possible to differentiate between liquid and vapor, as vapor generates a high-frequency hissing sound, while liquid does not. In addition, it allows easy detection of blocked valves due to the absence of high-frequency sound signals. Unfortunately, the data we had access to was from steam tanks monitored during the two-day test period and were newly installed, and no problems had been detected during the testing window provided by the facility. However, the device successfully demonstrated, in line with the findings of the literature, its ability to monitor both temperature and ultrasonic signals for this specific application.

http://Carles Ferrer Ramis