Smart Valves "Listen" to Machinery, Automatically Diagnose Issues
A strategic partnership between German valve OEM, SAMSON AG and artificial intelligence company, 3DSignals produces smart valves that use airborne ultrasonic sound data to not only sense changes in flow through them, but also to detect the health of operations in machinery around them.
3DSignals specializes in deep-learning neural networks to continually monitor the health of industrial equipment. They use ultrasonic microphones to record airborne sounds, and then send audio data to the Cloud for diagnoses of any arising issues by a deep neural network. Since it monitors airborne sound waves, the system can receive signals from several pieces of machinery, and then diagnose problems specific to each component based on the sound frequency.
3DSignals worked with SAMSON AG to engineer hardware that would integrate in the valve design, and survive harsh environments and contact with several media. They then built custom deep-learning neural networks that would identify deviations in the sound profile as diagnostics that are specific to hydraulic machinery. The valves will be on the market by the fourth quarter of 2017.
When released, technicians will be able to use the Cloud-based predictive analysis software to receive alerts and view sound profiles through an app. The smart valves will enable real-time monitoring of machinery from remote locations, helping customers to remain competitive in the age of Industry 4.0.