Hear What You've Been Missing
The simple act of listening can reveal a lot about our surroundings. With SensiML acoustic event detection Knowledge Packs, you can now train your IoT devices to recognize and respond to specific sounds. While the possible applications are nearly limitless, below are just a few examples that have been implemented:- Pump blockage and cavitation
- Livestock health and welfare
- Motor and drivetrain faults
- Building and perimeter security alerts
SensiML Acoustic Event Detection
SensiML employs the latest cutting edge signal processing, feature extraction, and classification techniques to provide accurate sound recognition models in extremely compact code. Our acoustic recognition models can be as small as 40kB and typically under 256kB of total memory size.Audio-Based Pump State Monitoring
This demo illustrates how SensiML Knowledge Packs can detect even nuanced state changes in pumps, motors, and other mechanized equipment. This particular model achieves 98.8% accuracy across five trained states of a centrifugal water pump including:- Pump Off
- Pump On - Normal Flow Rate
- Pump On - Low Water Level
- Pump On - No Water Remaining
- Pump On - Obstructed Intake
SensiML offers complete Knowledge Pack development services allowing you to focus on your application. Our team can devise a customized project plan using your application audio training data to create tailored recognition libraries ready to drop into your application.
For project teams with basic machine learning familiarity, desire to learn, or needing to undertake the tasks entirely in-house, SensiML offers its ML Analytics Toolkit suite. A true end-to-end workflow, SensiML Analytics Toolkit supports the complete process from data collection and labeling to Knowledge Pack generation and testing.
