New Roomba i7+ Has Persistent Maps, Selective Room Cleaning, and Automatic Dirt Disposal

iRobot's newest vacuum can clean just the rooms you want it to, and then offload its dirt bin into a special dock and keep on going

5 min read
iRobot Roomba i7
The Roomba i7+ has a special dustbin that integrates with iRobot's new automatic dirt disposal system.
Photo: iRobot

Today, iRobot is announcing the fanciest, most amazing, and most expensive addition to its line of Roomba robot vacuums. The Roomba i7+ can do two big things that no other Roomba can do: The navigation maps that it makes persist between cleaning sessions, meaning that the robot “remembers” the layout of your home and can use that data to clean in a smarter and more efficient way. And the i7+ can autonomously recharge on a new Clean Base dock that has its own large dustbin and vacuum system, and can suck the dirt out of the robot’s bin so that you don’t have to.

The Clean Base is cool enough that we’ve written a separate article about it that you can read here, but essentially, it’s a drive-on dock with an internal vacuum that can access the i7+’s dustbin through a special port underneath, and then transfer the contents of the dustbin into the dock’s own bin, which is large enough to hold about 30 Roombafuls.

While the Clean Base was a total surprise, what wasn’t a total surprise is that the i7+ now features persistent maps. Just like the previous generation of high-end Roombas, the i7+ uses a top-mounted camera that collects feature data about your home to generate a map through VSLAM (visual simultaneous localization and mapping). Earlier Roombas made a new map every cleaning cycle, but the maps that the i7+ makes persist between cleaning cycles—in other words, the robot remembers your floor layout and can plan accordingly the next time you ask it to clean.

Since the maps stick around, you can do useful stuff to them through iRobot’s app. That useful stuff is currently the ability to tell the robot where different rooms in your house are, and then direct it to clean only those rooms: “Robot, go clean the kitchen.”

Since the maps stick around, you can do useful stuff to them through iRobot’s app. That useful stuff is currently the ability to tell the robot where different rooms in your house are, and then direct it to clean only those rooms: “Robot, go clean the kitchen.” In practice, the robot will make an exploratory cleaning pass, and then show you a map of your place with suggestions about how it thinks your rooms should be divided up. You can then adjust (add, remove, reposition) the delineations and give the rooms whatever labels you like. Each robot can remember maps for up to 10 different floors, which is great if you live in a 10-floor mansion (like I don’t), and the VSLAM system makes it particularly resilient to normal environmental changes like shifting furniture.

iRobot Roomba i7 app The Roomba i7+ learns and remembers a home’s floor plan, so you can schedule it to clean specific rooms by name. Photo: iRobot

Now, not to brag or anything, but we totally saw this coming, because at IROS 2017 in Vancouver, iRobot researchers presented a paper on RoomSeg, an algorithm that (you guessed it) automatically segments rooms out of a robot-created map. In addition to enabling the cleaning of only specific rooms, the paper also included data on exactly how much of a difference the pre-existing map combined with room segmentation had on the efficiency of a cleaning robot:

While conventional systematic cleaning needed in total 2.32 hours, of which 0.52 hours were for path following and 1.8 hours for cleaning, room-by-room cleaning needed only 1.9 hours of which 0.3 hours were for path following and 1.6 hours for cleaning. Overall, the total execution time decreased to 82 percent, path following to 63 percent, and turning down to 66 percent. 

The i7+ also features a bunch of little improvements as well, or maybe not so little—one in particular that I personally appreciate is that the robot can now tell the difference between black carpet and the endless screaming void. In previous Roomba generations, the infrared cliff sensors could get confused on IR-absorbing black carpet, causing the robot to avoid those areas and freak out if forced onto one. That problem has now been solved, we’re told, and there was much rejoicing and removing of electrical tape and tinfoil

iRobot Roomba i7 bin is rinseable The dustbin is now designed to be rinsable. Photo: iRobot

Efficiency is certainly important and all, but you may not notice it all that much if you let your Roomba do its job while you’re not around. The i7+ is the most autonomous Roomba ever, when combined with scheduling and the Clean Base, since it can operate for extended periods without you having to manage it. iRobot says your floors will be just consistently cleaner. As iRobot CEO Colin Angle says in the press release, “this robot delivers on the original vision we had when we embarked on the journey to build a vacuuming robot almost 20 years ago.”

At US $950 for the i7+ along with the Clean Base dock, this is a wicked expensive robot. If you want the mapping features but not the automatic dirt disposal, iRobot is selling the same model (called just i7) for $699. And if you want just a vacuum robot and none of the mapping and other bells and whistles, iRobot is also today announcing the Roomba e5, which costs $450. For people who live in a relatively small place, and have the minor amount of patience that it takes to manage a vacuum that won’t do any of the fancy navigation stuff and can’t empty itself, we’re obligated to point out (as we usually do) that your floors will likely come out nearly as clean—that extra $500 for the i7+ is (almost) all in features that are useful but not necessary, along with convenience. 

iRobot Roomba i7 Software upgrades will allow a Roomba to autonomously learn to clean specific places that get most dirty most frequently, such that you won’t have to tell it when or where to clean at all. Photo: iRobot

Having said that, we’ve been really excited about this whole persistent maps business for home robots, and the i7 is iRobot’s first model to leverage the capability. What we’re seeing now from iRobot (and also from Neato Robotics) is just the beginning, and these robots are going to get smarter and more useful very rapidly from now on. For example, the i7 (like other Roombas) is able to detect which parts of your floor are dirtiest by listening to the amount of stuff that it’s picking up. Since it’s now localizing and remembering maps, over time it should be able to build up a map of where dirt tends to accumulate in your home. At some point, it should be able to autonomously learn to clean the specific places that get most dirty most frequently, such that you won’t have to tell it when or where to clean at all—maybe it’ll go under the couch once every few weeks, but do a pass around your front door every day or two. And there’s really no reason I can think of that iRobot couldn’t implement that particular feature, like, tomorrow. 

Beyond obvious things like that, you only have to look to interviews we’ve done in the past with Colin Angle to see where things are headed, and the iRobot team told us that we can “expect a lot of interesting stuff in the short term—the i7 is built with the future in mind.” We’re not exactly sure how iRobot will be topping this, but we’re excited to find out.

[ iRobot ]

The Conversation (0)

How the U.S. Army Is Turning Robots Into Team Players

Engineers battle the limits of deep learning for battlefield bots

11 min read
Robot with threads near a fallen branch

RoMan, the Army Research Laboratory's robotic manipulator, considers the best way to grasp and move a tree branch at the Adelphi Laboratory Center, in Maryland.

Evan Ackerman

This article is part of our special report on AI, “The Great AI Reckoning.

"I should probably not be standing this close," I think to myself, as the robot slowly approaches a large tree branch on the floor in front of me. It's not the size of the branch that makes me nervous—it's that the robot is operating autonomously, and that while I know what it's supposed to do, I'm not entirely sure what it will do. If everything works the way the roboticists at the U.S. Army Research Laboratory (ARL) in Adelphi, Md., expect, the robot will identify the branch, grasp it, and drag it out of the way. These folks know what they're doing, but I've spent enough time around robots that I take a small step backwards anyway.

The robot, named RoMan, for Robotic Manipulator, is about the size of a large lawn mower, with a tracked base that helps it handle most kinds of terrain. At the front, it has a squat torso equipped with cameras and depth sensors, as well as a pair of arms that were harvested from a prototype disaster-response robot originally developed at NASA's Jet Propulsion Laboratory for a DARPA robotics competition. RoMan's job today is roadway clearing, a multistep task that ARL wants the robot to complete as autonomously as possible. Instead of instructing the robot to grasp specific objects in specific ways and move them to specific places, the operators tell RoMan to "go clear a path." It's then up to the robot to make all the decisions necessary to achieve that objective.

Keep Reading ↓ Show less