Here’s a scenario for you: It’s been a while since you’ve done your laundry; the stuff is piling up and threatening to take over. So, reluctantly, you decide to take the bull by the horns—you load the machine.
However, despite knowing that you shouldn’t, you stuff the thing as full as you can get it. Just an extra towel won’t be a problem, right?
Well, as we all have experienced, the machine takes umbrage and starts thumping and knocking in the spin cycle.
That extra weight has unbalanced the contraption, and the thumping noise that you hear is actually telling you that there’s a problem with the machine.
Diagnostics
Start-up Augury is using a similar principal to diagnose machines in general.
It reckons that machine noise can be harnessed and used to diagnose problems. So the company is developing a mobile phone-based solution that automatically diagnoses machines based on the sounds they make.
Augury thinks sensors placed on machines can provide a heads-up on malfunctions if you know what the machine is supposed to sound like. Sound anomalies can indicate issues.
Sensors
Augury’s system connects ultrasonic and vibration sensors to smartphones, measures the sounds, and processes them with machine learning algorithms.
“Every mechanical system can be characterized by the sound that it makes. Machines ‘talk’ and we understand their language,” the company says.
Ultrasonic and vibration sensors are already used in predictive maintenance, but where this system takes it a step further is that it’s Internet- and algorithm-based.
The data is sent to servers where “it is compared with previous data collected from that machine, as well as data collected from similar machines,” the company says on its website.
The platform then detects changes.
“This analysis is done in real-time and the results are displayed on the technician’s smartphone within seconds,” it says.
Algorithms
The key to it all working properly is the algorithms. Recordings are stored on the company’s servers and should “get smarter and grow stronger” as the servers receive more information.
If the platform finds out about a new kind of malfunction, it studies it and adds it to the system.
The device
Augury uses a small, portable sampling device at the site of the equipment. An iOS app or web dashboard displays the results.
Sampling devices can be moved from machine-to-machine and the iOS app helps the technician with sensor placement. After placing the sensor, the technician can invoke the sound recording, after which it analyzes the recording and displays the results.
The app also suggests a treatment if it finds a problem. In my case, I would need more of a reprimand. Something like “remove the extra towel, and don’t be so lazy next time” would probably work.
Internet of Things
But, for our purposes, where this really gets interesting is in consumer-level IoT possibilities.
If smart appliances will be already connected to the Internet, it wouldn’t be that hard to add an acoustic element, coupled with the existing Augury algorithms.
That way, any IoT domestic appliance gains bonus preventative maintenance for not much additional work. It’s just one more sensor.