# Performance Optimization and Condition Management of Industrial Pumping System using Real-time Data Analytics

> With the support of the Savonia University Applied Sciences and Viimatech equipment, we implement a smart water project in the Savonia water laboratory.

## Overview
Optimizing the performance of industrial pumping systems is in increasing demand to save resources, reduce energy consumption, enhance productivity, and extend the life cycle of devices. By applying Industrial Internet of Things (IIoT) technologies, we can continuously monitor and optimize pump performance. This project, conducted at the Savonia water laboratory (WaterLAB), utilizes the Viimatech VT03 pumping monitor to record pump activities and analyze data to detect abnormalities and improve maintenance processes.

## WaterLAB and WaterLOOP Infrastructure
Savonia’s WaterLAB is equipped with WaterLOOP, a Distribution System Simulator (DSS) that operates 24/7. 
*   **Facility Details:** Nearly 1 km of pipework in a looped-branched network.
*   **Components:** Includes water intake, first and second-stage pumping stations, and a dual wastewater system with a separator tank.
*   **Capabilities:** Operates in closed or open-loop modes; equipped with specialized online sensors for real-time water chemistry fingerprinting.
*   **Research Utility:** The system is fully sandboxed, allowing for the simulation of contamination events, hydraulic disturbances, and leakages. It provides metadata for training and validating artificial intelligence applications.

## The SuperDMA Demonstrator
To bridge the gap between development and real-world application, Savonia offers the SuperDMA full-scale demonstrator. Located within the Savilahti Innovation Ecosystem in Kuopio, this facility serves as a model for a smart city, allowing for the testing of prototypes in a real-world water distribution system without disrupting normal operations.

## VT03 Sensor Technology and Installation
The project utilizes the VT03 device from Viimatech, which employs non-invasive split-core current transformers to measure pump current and power.

### Installation Steps
1.  Connect 24VDC power input to terminal X3 POW.
2.  Attach current transformers to motor phase conductors.
3.  Connect remaining inputs to correct terminals.
4.  Attach VT03 to DIN rail.
5.  Connect power cable and monitor LED status (Red: power on; Blue: connection starting; Green: fully operational).

### Data Analysis
The sensor monitors current phases and loops to evaluate four key indexes:
*   Running cycle length
*   Runtime average phase current
*   Time between cycles
*   Motor duty cycles

## Results and Findings
During a four-week observation period, the system successfully identified operational trends and potential failure indicators:
*   **Anomaly Detection:** Sudden current drops indicate suction line blockages or pin shear, while sustained high current suggests the need for lubrication or bearing maintenance.
*   **Usage Patterns:** Data analysis revealed peak activity between 21:00 and 23:00, with lowest activity between 03:00 and 07:00.
*   **Predictive Modeling:** A Long Short-Term Memory (LSTM) neural network was trained on 30 days of data to predict pump usage for subsequent hours.
*   **Correlation:** A correlation matrix analysis showed a weak linear relationship (0.2–0.3) between daily usage and other provided variables, suggesting usage is independent of those specific properties.

## Conclusion
The application of the VT03 device provides a fundamental base for digital twin development and machine learning-driven efficiency optimization. Future improvements may include the application of the MQTT protocol to enhance sensor system performance.

## Acknowledgments
We acknowledge the support of WaterLAB’s research manager, engineers, and Viimatech for providing devices for testing.

## Authors and Affiliation
*   **Dung N Lai**, IoT student, Savonia University of Applied Sciences
*   **Sameer Karn**, IoT student, Savonia University of Applied Sciences
*   **Patryk Wójtowicz**, Research Manager, Savonia University of Applied Sciences
*   **Rajeev Kanth**, Senior Lecturer, Savonia University of Applied Sciences