They may not come with flagons of brandy attached to their necks, but drones could nonetheless put St. Bernards in the Swiss Alps out of search-and-rescue work for good.
Researchers from the Dalle Molle Institute for Artificial Intelligence and the University of Zurich have designed a quadcopter than can respond to the 1,000-plus distress calls that emergency responders receive each year by using an A.I. algorithm to spot and navigate man-made trails.
By relying on a bank of more than 20,000 trail images captured by GoPro cameras worn by hikers, the drones were able to determine the direction of the trails 85 percent of the time, a three percent improvement on what human test subjects managed during trials.
The dogs, who have been trained to track down lost and injured hikers along Switzerland’s St. Bernard Pass since the 18th century, have largely been replaced by helicopters, though drones might be poised to take over operations in the area on a permanent basis.
“Now we that our drones have learned to recognize and follow forest trails, our next step will be to teach them to recognize humans,” says University of Zurich Professor Davide Scaramuzza.
That piece will be key if the designers ever want their “small flying robots” to come close to what Barry, a St. Bernard whose preserved body is on display at the Natural History Museum in Berne because he saved the lives of more than 40 people, was able to accomplish.
We reached out to a St. Bernard for comment but were unable to make any sense whatsoever of his response.