Connected gadgets in your home have grown increasingly connected, but healthcare is only just beginning to join this brigade. One of the biggest reasons why health professionals are reticent to connect their devices is a concern over security — for health tech, attacking a device can mean attacking a person.
Hospitals have sought out ways to mitigate digital attacks, however, and legacy antivirus programs have been the mostly common route. But researchers have recently been tinkering with a new method for detecting nefarious digital actions: a program called WattsUpDoc.
WattsUpDoc is a program that uses power and electricity as a means to detect if a malware has been introduced into a network. By analysing the energy used, researchers are able to accurately identify if a malware is in the network.
Hospitals are now slowly beginning to sign up and try out this new malware detection system. In fact, two “large US hospitals” have signed up to use WattsUpDoc to detect network vulnerabilities, the Register reports, though the specific names of the participating hospitals are still under wraps.
Interestingly enough, WattsUpDoc works by listening to the power outlets of these connected medical devices and looking for anomalies in power flow. “Flash memory actually draws power differently when it’s starting to become run down,” explained Kevin Fu, one of the professors behind WattsUpDoc who also performs medical device research at the University of Michigan.
Fu and his team have spent the past many years studying how these power drains differ and what they mean. Using this data, WattsUpDoc analyses computers’ energy, day in and day out, and “gives predictions with certain confidence.”
The thesis, which is now beginning to be proven, is that if a new actor is introduced into a network — namely, a piece of malware — something in the power flow will change. Tests claim that the WattsUpDoc Platform has been at least 94% accurate in detecting malware, which is basically the same accuracy that antivirus software currently has.
Fu also explained the genesis of his program. One of his students approached him and said “I can tell what website you’re browsing based on you power outlet.” Fu was amused by the idea — “it made for a fun cocktail party trick” — but he didn’t see a real use for it.
But Fu said he did see promise, so he decided to use that same idea but turn it upside-down. “Instead of looking for just what website you’re browsing,” the professor said, maybe he could “look for anomalies that tend to be bad things.”
WattsUpDoc was first presented to the public in 2013, but has yet to be officially tested in formal medical facilities. The researchers have been tinkering with it and improving its detection accuracy for years. “The secret sauce is a whole lot of machine learning,” Fu explained. “We know what good behaviour looks like.” Using this technology the platform analyses if there is a vulnerability in the system and how urgent of a problem it is.
Ever since it was first announced, Fu said medical professionals have been asking when they can try it out. Finally, at this year’s RSA conference, the team announced that two hospitals will begin trialling to platform.
If it proves successful, the team will roll it out to more medical providers.
Fu believes this could be the beginning of a new paradigm for understanding digital vulnerabilities. Up until now security has been thought of in a very black and white manner; Is there a malware or is there not? The WattsUpDoc team think it’s more nuanced than that. Using this technology, WattsUpDoc looks to answer “how well are you,” said Fu.
Looking at malwares from this angle could also change how other security professionals approach their work. “We’re trying to change the way people think about anomalies,” said Fu.
NOW WATCH: Tech Insider videos
Business Insider Emails & Alerts
Site highlights each day to your inbox.