[ad_1]
iRobot recently released a new robot vacuum that uses AI to prevent dog poop, but the company’s Roombas have long struggled with a more basic problem: confusing dark patterns on rugs and rugs with perilous falls.
A video recently shared on Twitter by IBM researcher Dmitry Krotov shows exactly the problem: A Roomba rolling inside a rectangular box pattern on a carpet, refusing to roll over the lines as if they were physical walls. Usually iRobot’s software reports this kind of problem as a dangerous “cliff” and a quick Twitter search for “roomba carpet cliff”Shows that this is a fairly common mistake. As a single user Complains: “I swear that Robby the Roomba is a true drama queen !! This is NOT a Robby Cliff. It’s a carpet!
The actual mechanics behind this error are quite fascinating, however. And it has nothing to do with machine learning, but rather the cost constraints on the hardware of robot vacuums.
As iRobot Research Scientist Ben Kehoe explained in response to Krotov’s video, the basic problem is that Roombas detects sudden falls like stairs and steps using a combination of an LED and a photodiode – a sensor that detects light. Like Kehoe the dish: “Does the photodiode detect the light reflected by the LED? Great, the ground is there! No reflected light? Uh oh, that must be a cliff. Dark black carpet -> no reflected light, it looks like a cliff!
The problem, Kehoe notes, is balancing the accuracy of Roomba’s sensors while keeping unit costs low. “Our newer models don’t suffer from this, but it took YEARS to figure out how to make the sensor robust to this while still being cost effective,” he said. tweeted.
[ad_2]
Source link