This Gadget Can Tell What’s Wrong with Your Air Conditioner by Listening to It
Augury’s gadget and iPhone app pay attention to ultrasonic sounds and vibrations to figure out what’s wrong with air conditioners and other big machines.
If you’ve ever tried to diagnose a car problem just by listening to it run, you’ve got a sense of what a startup called Augury is doing. Only instead of having human ears pay attention, it’s analyzing vibrations and ultrasonic sounds to figure out what’s ailing bulky machines.
The company, based in New York and Haifa, Israel, uses an iPhone app, gadget, and sensor to record motors and pumps, and then compares the data to existing machine recordings to determine whether or not the one being analyzed is working properly, or what the issue is if it’s not.
For now, Augury is concentrating on diagnoses for commercial buildings’ heating, ventilation, and cooling systems; CEO and cofounder Saar Yoskovitz says Augury can point out issues like a bearing that’s ripe for replacement or a need for lubrication or realignment simply by listening in.
The company believes that its technology can help technicians and also cut down on building owners’ maintenance costs over time. And since most malfunctions don’t happen overnight, Augury says it can also let you know before an existing issue becomes an expensive thing to fix.
“We can predict when it’s going to be critical, and when you’ll need to start thinking of repairing it,” says Yoskovitz.
Yoskovitz says Augury is currently working with service companies whose technicians are responsible for making repairs to buildings’ HVAC systems, as well as with some facility management companies.
To diagnose a problem with Augury, a technician attaches a magnetic sensor to the body of a machine—a pump, fan, or central air-conditioning unit for now, since that’s what Augury is attuned to. The sensor is physically connected to a tool Augury calls the Auguscope, which collects vibration and ultrasonic sensor data and sends that information to a connected iPhone; an Augury app on the phone punts the data to the cloud, where it’s analyzed on Augury’s servers and compared with other recordings made in the past of that machine and other similar machines. Then the app can give the technician a diagnosis of the issue.
Technicians can also move the sensor to different parts of the machine to record in different spots, in order to help pinpoint where a problem is located; Yoskovitz says that in a test of this, Augury was able to diagnose and find the location of a crack in a motor’s rotor bar, which led to a fix that cost a fraction of the cost of replacing the motor itself.
Eventually, he hopes Augury’s capabilities will be built right into consumer appliances, making it possible for your washing machine or refrigerator to let you know when a part needs to be replaced.
Chris Mechefske, a professor at Queen’s University in Ontario who studies vibration-based machine conditioning and fault diagnosis, says that the basic idea behind Augury isn’t unique. But its technique of comparing new recordings to existing ones could be helpful if the comparisons are made to the right machines—that is, ones that aren’t just similar but are also operating in very similar operating conditions.
If so, he says, “that gives you a real leg up in terms of picking out not just when you have a fault or some deterioration occurring, but what exactly that is.”
Keep up with the latest in machine learning at EmTech MIT.
Discover where tech, business, and culture converge.
September 11-14, 2018
MIT Media Lab