Data acquisition module
Data science building blocks for machine data
SSV has developed a data acquisition module for machine sensors plus software to simplify the use of AI in machines, systems and industrial processes.
Data science aims to improve decision-making by extracting findings from large volumes of data and using them as additional knowledge for decision-making. Algorithms from the field of artificial intelligence (AI) are used for this purpose. In automation, providing suitable data is the biggest challenge before AI algorithms can even be used. The IO/5640-DS module and the PyDSlog Python software library were developed to solve this specific sub-task.
On the input side, an IO/5640-DS has eight analogue channels for digitizing sensor data, which are combined into a constant data stream. The raw sensor data is transferred via a 2-wire high-speed connection either to a PC using a USB adapter or directly to an edge gateway. The number of channels, sampling rates of up to 435 microseconds (2.3 kHz) with 12-bit resolution and communication block sizes can be set to suit individual requirements.
The PyDSlog library enables the acquisition of so-called "labeled training data", from which the necessary models for the practical use of machine learning algorithms and artificial neural networks can be generated. Various sensor data such as current, voltage, vibration, microphone level, etc. can be used as input. The output is a CSV file that can be used directly for training the respective algorithms or for manual data analysis under Python, R or Matlab.
Furthermore, SSV offers all IO/5640-DS and PyDSlog users a webinar with the following content: 1. basic principles and terminology of machine learning. 2. a complete machine learning process, including sensor data acquisition, data preparation, modeling and model evaluation 3 Determine model accuracy and adjust hyperparameters. 4. connect the output of a machine learning algorithm to other systems.









