Machine Learning and Sensor Data Analytics

Document created by edk Employee on Jul 7, 2016Last modified by Diana Torres on Apr 5, 2017
Version 6Show Document
  • View in full screen mode

Smart Pump Monitor Demo

This demo shows a small water pump rig consisting of a water pump and 3 valves put together to collect data for supervised machine learning. Normal operation as well as abnormal conditions may be simulated with the rig. There are 2 sensor boards attached by clamps to the water pipe. Each sensor board has many sensors on it, but only the accelerometer will be used to gather the data. One board is used for data logging.  The other runs a model which was generated via machine learning based on data logged from the first board.  Pump vibration measurements are processed through the model by the MCU on that board to determine the operating state of the system

 

 

Features

  • Use of accelerometer to measure pipe vibration
  • Sensing algorithm detects when the pump is clogged or drawing on air
  • How to find patterns in data taken by NXP Sensors

 

Links

 

Related demos

 

 

1 person found this helpful

Attachments

    Outcomes