Application Design for anomaly detection of 4.0 machinery
Risk Prevention Engine
Provision of an advanced analytics tool that allows the customer to anticipate potential risks and anomalies that could occur on 4.0 machinery with the aim of improving the maintenance and setup phases of the machinery itself.
Description of activities
Data ingestion phase managed via native APIs of 4.0 machines with consequent development of a log management pipeline with ElasticSearch indexes regarding anomalies.
Development of an anomaly detection model based on historical series of machine logs related to alarms mapped with different levels of danger.
Using Kibana to analyze the data collected in the form of logs and developing the end-user interface web app to interact with the anomaly detection model.
Tools and technologies used
ElasticSearch, Python, Streamlit, Prophet
