Sensor Forecasting
Why this use case matters
If applied properly, time series classification can be a powerful tool in extracting valuable business and operational insights from dense data streams. In the case of industrial machinery, and really any sensor information, the machinery heath-state can be classified from sensors that are continuously measuring power, temperature, pressure, vibration, etc.
Techniques
Time-series classification (TSC) models
Algorithms
Distance-based KNN, TimeSeriesForest, Boss, cBoss, RISE, Recurrent Neural Nets (RNN), LSTM
Outcome
Predictive & prescriptive insights extracted from high velocity sensor streams.
Mosaic worked with a leading animal technology company to use machines to provide pet & animal owners with more visibility into the quality of life their pets have.
Mosaic has compiled our industry expertise into a Machine Learning playbook for Manufacturing.
