Challenge
Early detection of black mold, which grows in damp and humid areas and causes extensive damages to common construction materials as fiberboard and drywall.
Solution
Data captured by humidity, temperature, and water leakage sensors was used to develop algorithms that detect the risk-level of black mold growth.
Challenge
Complete automation of molten metal pouring in foundries with data &
algorithms removing any dependencies on the operator.
Solution
Machine learning algorithms with PLC simulation data for training are used to
automate the metal pouring operation. The solution is being deployed in multiple foundries globally to pour molten metal automatically in real-time.
Challenge
Reliable anomaly detection with low false alarm & troubleshooting for a critical air-blower that feeds combustion air to major parts in the refinery.
Solution
A semi-supervised anomaly detection solution based on expert-annotations provides alerts in real-time for deviations from normal behavior. A cause for the deviation is also indicated.
Challenge
Automatically optimize & control the energy consumed by electrolyzers at a chlorine plant based on their predicted health, the energy price variations in the R2 imbalance market & the production targets.
Solution
A predictive model for the health of the electrolyzers was developed & the energy consumption was optimized based on mathematical & data driven models. Further a mathematical optimization was performed in order to account for fluctuations in energy prices, customer orders & storage.