AquaNERVA: The Next Big Thing For Municipal Waterworks?
AquaNERVA is a centralised control and monitoring system for the management of treatment processes relating to water and wastewater. The use of data analytics and machine learning techniques has made this an intelligent tool for condition monitoring, predictive diagnostics and prognostics — and one of the smartest water management systems in the world.
Designed and developed by the Marine arm of ST Engineering in 2018, AquaNERVA is currently being deployed at the Singapore Public Utilities Board’s Ulu Pandan Water Reclamation Plant (UPWRP). According to Mr Sim Chee Chong, a technology and solutions expert at ST Engineering Marine, AquaNERVA could be the answer to enhanced water quality, increased water security, and improved operational efficiency for municipal waterworks.
Here’s how UPWRP has been benefitting from AquaNERVA:
Prolongs Maintenance Intervals
Maintenance is performed based on condition of the equipment as opposed to a fixed maintenance schedule. This improves uptime and extends the lifespan of equipment, hence generating cost-savings.
Reduce Unplanned Downtime
The use data analytics and machine learning prevents unplanned downtime and enables predictive diagnostics, which are essential features for uninterrupted operations. This changes the nature of maintenance from reactive to preventive. As the smart system is able to detect and evaluate anomalies, alert operators to potential problems, as well as recommend actions for repair and maintenance, operational cost is greatly reduced. Thanks to the incorporation of machine learning techniques, the system is also able to learn from past equipment failures and predict future ones more effectively.
Mr Sim explains, “We’ve actually designed and deployed a similar system in our ships. We call it the Sensemaking system, but essentially, it works the same way by applying data analytics and machine-learning techniques to monitoring systems in large-scale engineering projects. We have seen positive results and the system has been very consistent in predicting potential failures.”
A common challenge in the implementation of such systems is the lack of quality data. “Massive amounts of good quality data are needed to train the machine-learning model, so as to ensure the model is intelligent enough to perform its tasks. We overcame these challenges by working closely with our customers, so that we can identify the areas of improvement in data management, and provide follow-up advice,” Mr Sim assured.
For sales enquiries and demonstrations, please contact Mr Ng Tee Guan at NG.TeeGuan@stengg.com