iRobot just lately launched a brand new robotic vacuum cleaner that makes use of AI to keep away from canine poop, however the firm’s Roombas have lengthy struggled with a extra primary downside: complicated darkish patterns on carpets and rugs for perilous drops.

A video recently shared on Twitter by IBM researcher Dmitry Krotov exhibits precisely the issue: a Roomba rolling round inside an oblong field sample on a rug, refusing to drive over the traces as if had been bodily partitions. Normally, iRobot’s software program flags this form of downside as a harmful “cliff,” and a fast search on Twitter for “roomba rug cliff” exhibits it’s a fairly frequent error. As one person complains: “I swear that Robby the Roomba is a right drama queen!! This is NOT a cliff Robby. It’s a rug!”

The precise mechanics behind this error are fairly fascinating, although. And it has nothing to do with machine studying, however fairly the price constraints on the robotic vacuums’ {hardware}.

As iRobot analysis scientist Ben Kenhoe explained in response to Krotov’s video, the fundamental downside is that Roombas detect sudden drops like stairs and steps utilizing a mixture of an LED and a photodiode — a sensor that detects gentle. As Kenhoe puts it: “Does the photodiode detect reflected light from the LED? Great, the floor is there! No reflected light? Uh oh, that must be a cliff. Dark black carpet->no reflected light, it looks like a cliff!”

The issue, notes Kenhoe, is balancing the accuracy of the Roomba’s sensors whereas conserving unit prices 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 remaining cost effective,” he tweeted.


Please enter your comment!
Please enter your name here