OSCA: wastewater transport diagnostics using big data, AI and ML

Water authorities monitor water transport systems on a daily basis, and in recent times the number and complexity of pumping stations in catchment areas has increased. Despite this, the sector has seen a scaling down of staff numbers and expertise. To offer a helping hand, Witteveen+Bos has developed the OSCA diagnostic tool.

Continuous monitoring

OSCA stands for ‘optimal sewage water system control AI’. It is a diagnostic tool that provides information on handover levels (‘afnameverplichtingen’), the existence of any leaks in the pipe network and the condition of pumps. OSCA combines big (weather) data with sensor data and continuously calculates how a water system is performing in terms of predefined indicators and variables.

Indicators used by OSCA include handover levels, pipe condition (resistance) and pump condition (using pump curves). By continuously monitoring these, it can detect anomalies – for example, due to leaks, blockages or the end of pumps’ lifespans – fast. The indicators are calculated using hydrological and machine learning (AI) models, resulting in data that can be displayed on a dashboard within 15 minutes. In the event that an anomaly is detected, OSCA immediately alerts the user.

For operators and experts

OSCA was developed in collaboration with Waterbedrijf Limburg and is tailor-made to meet daily-practice needs, offering a range of possibilities for satisfying the wishes and requirements of both process operators and hydrology experts. A user-friendly dashboard displays steering information relevant for both groups:

  • For operators, OSCA calculates how often a water authority has satisfied (actual) and will satisfy (predicted) its handover levels (RWA/DWA);
  • For hydrology experts, OSCA makes it possible to run automated analyses on pumping stations. This could help determine, for example, what would happen if heavy rainfall occurred after an area had been suffering from extreme drought for several days. By using both the theoretical model WANDA – or ‘digital twin’ – and machine learning models, OSCA can provide a good indication of what the impact of this would be. It can also determine whether the transport system will satisfy the required handover level.

OSCA operates entirely in the cloud and has been successfully field tested. Its advantages include scalability (catchment areas and/or pumping stations can easily be added), reliability and security.

More information?