To address the Industry 4.0 initiative to make “Smart Factories”, Predictive Maintenance has been a key area where investment is made to monitor machines, pumps, motors, etc.
Here is a high level application for the monitoring of these machines where the HW components are part of a sensor module that is attached to the machine that is being monitored.
The sensor module has an MCU that is typically combined with BT/BLE or Wifi or other connectivity along with sensors that are ideal for detecting different performance issues or anomalies.
One sensor data ML model doesn’t fit all machines as they are mounted on different surface (rubber, concrete, wood). Collecting data onsite and building models quickly and efficiently will allow each machine to have its own unique AI model that runs on the “endpoint” and “edge”. Qeexo AutoML is an end-to-end Platform that can create these unique models quickly and efficiently.
As low power MCU and low power connectivity has become more prevalent, so has the growth of structural health monitoring. In Japan for example, there are many tremors and earthquakes regularly which are hurting the fatigue of bridges, railroad tracks, roads, and buildings.
The Japanese government is investing in putting these small sensor modules on the bridges, buildings, etc. to monitor the “health” of the structure.
These sensor modules would consist of high precision, low drift/bias accelerometers as well as a low power MCU and a low data bandwidth connectivity technology like LoRa.
These accelerometer ML algorithms will require machine learning, that is easily delivered by the Qeexo AutoML Platform.
In smart home appliances, the microphone has proven to be an important sensor to make these devices matter.
Some use cases for a microphone that can benefit smart home appliances include:
All these uses cases need machine learning algorithms to make sense of the sensor data. Qeexo AutoML can automate this process for OEMs.
Qeexo AutoML can automate the creation of machine learning sensor algorithms for the following wearable and IoT applications:
All of these applications have the same architecture in common: An MCU with connectivity and sensors. Qeexo AutoML can create the ML algorithms quickly and efficiently that can run on these small battery powered devices.