Apply AI on any data without expert knowledge
- Drag-and-Drop import of excel files
- Browse and select data sets within your DBs
- Organize your data sets in separated work spaces
Knowledge-based data preparation
- Evaluate the validity either for the whole data set or for individual signals based on automatically imported min/max ranges and quality codes
- Apply your expert knowledge and adapt min/max ranges or reaction times in order to improve the automatized data cleaning
- Visualize dependencies between signals in order to learn more about your data sets
Guided training phase for predictive modeling via AutoML
- Train a model without advanced knowledge or specialized skills.
- Let AIXPERT choose the best algorithms, models or combinations thereof, depending on the nature of problems or use cases
- Achieve reliable predictions using the latest advances in RNN, LSTM and more
Learning from failures by using pattern recognition
- Manually mark abnormalities in sensor signals
- Test and evaluate the pattern detection directly on historical data and find the same phenomenon in the past
- Export the pattern as a black box and integrate it into RULESXPERT
- Combine different patterns to detect emerging equipment failures
Export models as „black-box“ for their integration in RULESXPERT
- Import and evaluate data-driven models in RULESXPERT
- Integrate AIXpert models as well as advanced models as R and Phython code
- Run your ML model directly in operation (on the edge) or securely hosted in the cloud
« Existing process models are the result of many years of experience of our process engineers. They are not to be replaced. In order to integrate new data-driven approaches, our process engineers can now independently and intuitively train their own AI models. Machine Learning is successfully accepted not as a job killer but as active support for their work. »