Early warning of filter breach
Use case
Challenge
Filtering is a critical step in the food and beverage industry, preventing food spoiling, prolonging equipment lifetime, and in some cases giving a better taste to the end product. Filters are prone to clogging and breakage and therefore require regular maintenance.
An Intelecy client in the food and beverage industry uses filters during production to remove bacteria spores resistant to pasteurization. There have been cases where filter breaches have allowed bacteria spores to pass through the filters spoiling the final product. As the final product is tested via intermittent laboratory testing, this can lead to a considerable loss for the company. Filter breakage is challenging to detect as it results in minor pressure changes.
Solution
By creating anomaly detection no-code AI models with Intelecy to monitor the filters, the client now receives early warnings of filter breaches. Multiple models were built and deployed on multiple filters tuned according to earlier known cases, allowing the factory operations team to identify production issues as they occur, rather than after it’s too late. This actionable insight helped to improve their efficiency, reduce waste and save money.
Result
- Wastage was reduced
- Unplanned downtime was avoided
- 2.2m USD annual savings
Other Food & Beverage use cases
Forecasting wastewater
By using Intelecy no-code AI, you can see the future of production processes and, for example, predict when it is the best time to take action. Learn how process engineers use no-code AI models to predict the future temperature and pH values and thus be able to minimize environmental impact.
Quality related anomaly detection
High quality of end products is crucial. At the same time, process engineers want to prevent unwanted incidents in production that lead to raw materials being lost as waste. Read more about how to gain stable quality by letting no-code machine learning models monitor each subprocess in real-time to detect deviations and prevent waste.