A semiconductor giant offering a wide range of application processors and MCUs with secure connectivity for industries like automotive, industrial & IoT, mobile, and communication infrastructure markets.
Client wanted to deliver a solution for condition monitoring of various industrial machines, HVAC systems, smart home appliances, etc. based on early detection of anomalies using client’ edge platform to allow preventive maintenance and avoid equipment damage. For the same, client was looking for a development partner with expertise in Machine Learning.
VOLANSYS is working as client’s development partner with hands-on experience in Machine Learning to enable health monitoring and predictive analysis of home appliances, HVAC systems and industrial machines due to anomalies detected using edge analytics based on Machine Learning. The model developed detects the anomalies like vibration, temperature, power, etc. and monitors the condition of the devices.
- Machine Learning
- Supervised Machine Learning on structured data captured using 3-axis accelerometer and gyroscope
- Developed Python script for time and frequency analysis, signal processing, feature extraction of vibration data captured
- Developed Python model using TensorFlow and trained using AWS SageMaker
- Developed script to export model parameter and convert to TensorFlow Lite format to deploy on edge
- Developed application on NXP i.MXRT platform to collect vibration data and detect anomaly
- Model performance testing using tools like TensorBoard, What-If, ML Perf, TensorFlow Playground, etc.
Python | TensorFlow | AWS SageMaker | TensorFlow Lite | TensorBoard | What-If | MLPerf | Supervised Machine Learning
- Enabled client to capture new market with application developed on new processor series
- The solution delivered increased asset reliability, productivity and determined future risks well in advance