INA – Intelligent Anomaly Detection for miniaturized sensor systems and data loggers
The aim of the subproject "piSmart-Ina" is the recognition and visualization of anomalies based on heterogeneous sensor data collected with intelligent data loggers. This involves analyzing data with pattern recognition methods and porting anomaly classifiers from analysis platforms to the data loggers so that they are independent and autarkic.
Companies and Developers.
For the development of machines or condition monitoring of machines, animals, or humans, it makes sense to obtain and understand the current situation or operating conditions. For this reason, data recording systems are available in the market, which can store one or more sensor values. By collecting different signals, a new quality of information provision is possible by incorporating machine learning methods. While up to now sensor data has been stored by data logger and subsequently analyzed extensively by experts, the project is developing an autonomous system for automated pattern and anomaly detection from stored sensor data.