The Best Robot Vacuum
The best robot vacuum for most people is the $400 iRobot Roomba 650. It’s easy to use, works on bare floors and carpets alike, and leaves out the unnecessary features that drive up the price on most bots. And when something goes wrong (an inevitability over time) a fully modular and easy-to-service design means you can repair rather than replace.
If you schedule Roomba to run a few times per week, you won’t have to think very often about keeping your floors tidy. While it won’t fully replace your regular vacuum, it will prolong the time you can go between manual cleanings with very little effort and oversight on your part. Indeed, almost every household can benefit from a robot vacuum—it’s just a matter of whether or not you’re willing to pay for it.
No carpets or pets? Check out the $200 iRobot Braava 320 floor cleaner instead.
How do I know? Over the course of 22 hours researching the category, I looked at about 50 different robotic vacuums and floor cleaners. Based on conversations with editors from CNET and Geek.com, moderators at the RobotReviews.com forums, and an allergy expert, as well as info from tons of expert and user ratings, I whittled the list down to a few semifinalists. Then, I lived with the cream of the crop for a few weeks to get a feel for how those actually work.
Will a robot work in my house?
Yep, bots will be helpful in most houses and apartments. They work well on medium and short-pile carpets, as well as pretty much any bare flooring surface. They can steer around furniture (and often under it, since most bots only need about three inches of clearance) and are pretty good at stopping themselves from getting tangled.
Most of the reasonably-priced models we considered have a maximum cleaning area of 1,200 square feet, though 800 square feet seems to be more typical, judging by user reviews. The real limit has more to do with how much debris is on the ground and the openness of the floor plan. If you have lots of pets and lots of furniture, temper your expectations for how much ground a bot is able to clean in a cycle. Pet hair can stuff up the bin, leaving no room to pick up other debris that cycle. If the bot is constantly bumping into chairs and tables, it spends more time doing corrective steering and less time actually cleaning. Every so often, the bot will find a way to get stuck between chair legs or under the couch. But even with the hairiest dogs and most cluttered layout, robots will still reliably clean a few hundred square feet.
What if your house is bigger than the maximum cleaning area? The bots don’t remember your floor plan once a session is finished, so there’s no disadvantage to just picking it up and starting a cycle in a different room or floor. The good models all have ledge sensors, so you shouldn’t have to worry about your bot tumbling down the stairs.
That said, an hour of automated cleaning does more than 10 minutes of half-assed manual vacuuming.
“[Robots] are best at what I’d call maintenance cleaning,” says Sal Cangeloso, who has reviewed most of the iRobot Roomba vacuums for Geek.com over the past five years. “The human does the big clean, say once a month, and then you have the robot clean a few times a week. This’ll keep your place clean and make it so that a few missed corners and stuck-on grime aren’t a big deal.”
Certain layouts and flooring types aren’t particularly friendly to robots, either. They aren’t equipped to handle elevation changes of more than a couple centimeters, so tall thresholds might as well be walls. High-pile carpets are problematic—the modest suction and small agitator can’t really clean anything that thick, and the bot bodies are heavy enough to sink into the soft surface. Area rugs with tassels around the edges are a tangling hazard, if the strings are longer than two or three inches. Most bots come with at least one “virtual wall,” so if you have a room that will pose problems, you can block it off and let the robot run free everywhere else.
How we picked
The original Roomba came out nearly 12 years ago, but very few serious competitors have emerged since then. iRobot, the company behind Roomba vacs and Braava floor cleaners, is still the heavy hitter. Their bots are consistently well-reviewed by experts and owners alike. And who could forget DJ Roomba?
The only noteworthy challenger is Neato, whose robovacs have earned praise for powerful cleaning capabilities and a straightforward navigation system. That’s really it. All of the other brands we investigated are poorly reviewed, prohibitively expensive, or both.
Since it was pretty easy to settle on iRobot and Neato as the two focal points, it was just a matter of settling on the right models within each lineup.
Since these are vacuums, cleaning capability was the top priority. In a group comparison, CNET gave 4-star “excellent” ratings to the Roomba 790 and Neato XV Signature Pro. (That’s while the LG Hom-bot Square only managed a “Very Good” and the Infinuvo CleanMate QQ5 received just an “OK.”) These are both high-end models, but most cheaper Roomba and Neato bots have the same vacuum components (motors and agitators) as their more-expensive siblings. So we set our sights on cheaper models that clean just as well as the expensive bots CNET tested.
Reasonably smart navigation is also important, but there’s no way to spot this on a spec sheet. A bot should have the wits to cover all the nooks and crannies of a room, avoid tangling itself on cords, and should get back to its dock under its own power (most of the time, at least). In the Roomba line, the cheapest bots have the same nav systems as the halo models, so the extra cash doesn’t buy you a smarter robot. With Neato, there’s no indication on any spec sheets or marketing literature that the bots use different navigation. But CNET noticed a difference in their testing, finding that the newer models seem to be “smarter” about navigating through crowded, cramped areas—even when the firmware on the old models had been updated. More on that later, if you’re interested. But the short version is that the differences will not make a big difference for most people.
We settled on two semifinalists: The $400 iRobot Roomba 650 and the $335 Neato XV-21. Their vacuum parts are the same as on more expensive models. They have scheduling. Their nav systems are effective. And they also leave out most of the feature-bloat: extra dirt sensors, full-bin indicators, and touchscreen controls—stuff that isn’t likely to be missed.
We didn’t design any specific tests for this, really, because there’s already good data at CNET about how these bots clean in controlled settings. The real advantage to be gained from going hands-on is seeing what it’s like to live with both of them, so I brought them into my home for a few weeks, and used them like any bot owner would use them.
Let’s start with the Roomba 650’s obvious advantage: it simply cleans better in real-world situations than the Neato XV-21. In the first few cleaning sessions, both bots would return with full dustbins—they were cleaning areas that I’d missed or been to lazy to clean properly with a big vacuum. But after a few weeks of regularly scheduled cleanups, the Roomba 650 still managed to find enough cat hair and debris on my floor to fill itself, whereas the Neato XV-21 started coming back with less and less in its bin.
CNET has some useful test data on how the Roomba and Neato each perform on floors versus carpets, and with different kinds of particles like rice, pet hair, and sawdust. In those controlled tests, they’re pretty closely matched, though the Neato emerged with a slight advantage because it was able to get much more pet hair out of the carpeting.
In practice, though, the Roomba has the advantage because its iAdapt navigation system lets it pass over the same spots a few times per cleaning session. It looks like it’s scooting around randomly, coasting until it bumps into a chair or a wall or a cabinet, then changing course and repeating until it gets tangled or lost or the battery gets low. But there’s a method to the madness. Cangeloso says that it’s “the smartest” of all bots. Rich Brown at CNET says that watching it work is “like a party.” Watch it long enough, and it’s clear that there’s some sort of logic, even if it’s difficult to discern. According to the Roomba manual, it starts a cycle in a sort of Fibonacci spiral to establish the boundaries of a room, and then works from there—once it figures out where the walls are, and then runs a specific cleaning pattern to clear dust away from the edges with its spinning side brush.
So if the Roomba it doesn’t quite get all of the grit out of the carpet on the first pass, it’ll pick up a little more by the third and fourth time it goes over the same spot. (Cycles last for a little over an hour before the Roomba starts heading back to the dock to recharge—that is, of course, when it doesn’t get stuck somewhere else first). Sometimes, it’ll just miss a spot in a cleaning session—I found one or two dust bunnies leftover after some cycles. But after it runs three or four times per week, it’ll have covered all the ground in its cleaning area.
Some owners feel very strongly that the Neato uses a better nav system. It uses lasers to map out a room before it starts cleaning, and then follows a very obvious, deliberate, linear pattern. It passes over each spot once (maybe twice), and doesn’t miss many patches.
But in our real-world experience in an apartment with a mix of flooring and carpeting, the Roomba just picks up more stuff than the Neato. It may not look very elegant if you watch it work, but for most floor plans, its persistence yields better results than Neato’s logic.
People who own it love it—at Amazon.com, it has a user rating of 4.6 stars, the highest of any bot that we scoped out and unusually high for any product on Amazon. Cangeloso told us that, as the cheapest Roomba with scheduling, it’s the bot that he’d buy if he wanted to replace his old 500-series model. CNET praised the Roomba line’s cleaning prowess, too: Reviewer Katie Pilkington wrote that it “was the top performer on hardwood in all but one of our tests and a close second in most of the carpet tests.”
The UI on the Roomba 650 is super simple. Setting the clock and scheduling cleanings is a lot like using an old alarm clock; there’s nothing wrong with that (though, c’mon, it’s a little silly that iRobot couldn’t include some sort of touchscreen on a $400 gadget). Owners can schedule one cleaning per day, and the calendar repeats itself every week. If you want to clean outside of the schedule, you can just press the “clean” button on the bot and send it on its way. It has a spot-cleaning feature for concentrated messes, though I usually just found it easier to use a dustbuster.
Day-to-day maintenance is a lot like any other bagless vacuum or dustbuster, too. Just pop out the dustbin and empty it into the garbage about once per week (or more often if you have pets). Shake the filter off when it looks dirty, wash it when it looks super-dirty. And keep an eye out for tangles around the rollers, like you should with any vacuum. Long human hair was a bigger problem than cat hair in my testing—any strands longer than a few inches will probably get wrapped up. (Check out CNET’s review of the Roomba 880 for an idea of how long-haired dogs can seriously jam things up.)
Easy repairs also help Roombas stand out from other bots. It’s a modular system, basically, so one broken component doesn’t ruin the whole bot. “The key is that almost all the parts are easy to replace,” Cangeloso says. “Since I’ve had [my 500-series Roomba], I’ve replaced almost all the parts, because it’s cheap, easy, and frankly fun to do so. The most expensive part is $45 but a lot of the time these things will be under warranty.”
Flaws but not dealbreakers
So we already know that the Roomba vacuum’s performance falls way short of any upright or canister cleaner, though uprights and canisters don’t have robots attached to them. Its nav system is imperfect, though it’s better than any other solutions. And it collects a serious amount of hair around its bearings and spinning parts, enough to cause a breakdown if left uncleaned. But it only takes a few minutes every couple of weeks to keep things moving freely.
While the Roomba’s adventurous, free-wheeling nav system helps it pick up more debris than the Neato, it also leads the bot to getting stuck, tangled, or lost more often. I ran the Roomba 10 times over the course of three weeks, and it only once did it successfully make it back to its dock under its own power. When I’d find the bot, it had usually sucked up the end of a rogue USB cable or charger and shut itself down to prevent damaging itself or the cord. One time, it ate the edge of a light area rug and couldn’t free itself (though it was able to steer up onto the rest of my area rugs pretty easily). On another occasion, it somehow found a laundry bag stashed under my bed and got the drawstring tangled around the side brush. The really mystifying one was when it wedged itself between the legs of my office chair and couldn’t get out.
Whatever the case, even the neatest people may still find the Roomba getting stuck occasionally. It’s safe to assume you’ll probably have to track the bot down and bring it back to its dock at least some of the time.
Other little gripes: The 650 comes with one virtual wall, which is an invisible wall that I sometimes used to keep the bot out of the bathroom. It’s handy. But it runs on 2 C batteries, not included. Who keeps C batteries? The good news is that the wall should run for 6 months per pair of batteries. The bot will also run right into your pet’s food and water bowls, probably pushing them around, definitely spilling water. And while the 650 isn’t as loud as other bots, the sound still gets obnoxious after a few minutes.
On the whole, the Roomba was worth having around—even actually kind of fun, like an extra pet that spent most of its time cleaning up after the other pet. The bleep-bloops are pleasant and friendly. I got a kick out of watching it steer its way into trouble and then (usually) finding its way out. It won’t clean everything, and needs some attention to work properly, but it’s as good as a bot gets at this price.
Long-term test notes
After nearly five months of regular usage, it’s been smooth sailing with the Roomba 650. Every other week, I need to cut tangled hairs away from the rollers, as I would with any vacuum. Once every four to six weeks or so, I pop out anything that spins, including the rollers, the front wheel, and the side brush, to clear the bunched-up cat hair. The original filter is still good after a few rounds of washing. It still runs great, even after some serious work while my cat shed her winter coat.
During the initial testing period, the Roomba had some trouble finding its way back to its dock. Now, the Roomba ends up back at its dock almost every time—but it’s not because the bot learned my floor plan, it’s because I learned how to clear the way for the bot.It’s become second nature to spend two minutes folding up the area rugs that give it trouble, and to stash the charging cables that it can choke on. I also moved the dock to a more central location, so that the bot is more likely to pass by the docking signal at the end of its cleaning cycle.
The bot’s body does look a bit banged up after five months of service, with some shallow scrapes around the sides from when it bonks into walls and chair legs. If I had nicer furniture, I might put a foam bumper around the bot’s bumper to prevent impact marks and smudges. But I’ve dinged up my apartment more with upright vacuums than anything that this bot has done.
No carpets or pets?
It won’t get the stains that are stuck on the hardest (like dried tomato sauce), and I don’t think it’s a great pick for people with several hairy pets, because hair covers the surface of the dry cloth pretty quickly. The navigation cube also runs on D batteries (included, thankfully). You can also use disposable cloths rather than the washable ones, but the cost adds up quickly and it’s just a wasteful practice.
The only other possible rival is iRobot’s Scooba line. These robot moppers are better at cleaning grime than the Braava bots, but they’re more expensive, too, and some owners have complained about their Scoobas breaking down after about a year. The Scooba 450 was announced at CES in January 2014—it could be a breath of fresh air for the line when it comes out this spring, but it’s $600, which is just way too much to pay for any single-purpose robot.
Why no step up or step down?
Below the sweet spot ($350-400), robots lose at least one of the three important features—vacuuming prowess, scheduling, or smart navigation. Check out CNET’s Infinuvo QQ5 review for an example of what you can expect from a cheap, white-label bot.
Spend more than $400, and you’re just paying for feature-bloat. Toss-ins like touch-sensitive controls, full-bin indicators, extra dirt sensors, and Wi-Fi scheduling do very little to improve the cleaning capabilities or user experience but drive the price up by hundreds of dollars.
Higher-quality filters get dangled out there as an upgrade. But we couldn’t find any mainstream robots that have earned a reputable air-quality certification. A representative from the Asthma and Allergy Foundation of America (AAFA) told us that there’s no way that any bot would earn their badge. A tighter HEPA filter could provide some incremental benefit but won’t magically transform a bot into a sealed, clean-air system.
Some of the extra navigation equipment can come in handy. iRobot’s “virtual lighthouses,” for instance, can help a Roomba distinguish individual rooms from each other. The bot will concentrate on thoroughly cleaning, say, your entire living room before wandering into the kitchen. The lighthouses are only compatible with higher-end Roomba models (we made an error earlier when we said that they worked with the 650), and only come bundled with the very top-end Roomba 880 ($700). Is it a game changer? Probably not. But it might help your bot perform more predictably in large homes with lots of rooms. We think you’ll be fine without a lighthouse, but if you want to try, the best value is to go with the Roomba 770 ($500), and add a lighthouse a la carte ($40).
In general, bots are more expensive than they should be. Mike Fortuna, a forum moderator at RobotReviews.com says, “my biggest disappointment since becoming a robotic vacuum user in 2005 is the prices have gone way up but the robots have not improved to match the price increase.” Our semifinalists are midrange robots, but cost more than most full-size vacuum cleaners (including some very good ones). Cangeloso says that the current generation of robovacs isn’t really an improvement over the older generation. You just get new designs and price tags that keep creeping upward. He notes in his Roomba 650 review that the bot’s firmware is an improvement over what the 500 series came with, but there aren’t many significant hardware changes. Then again, if it’s been more than a decade since the original robot vacuum and nobody has found a significantly better design, maybe the value of a proven bot counts for more than it seems.
Why not Neato?
Our other semifinalist, the Neato XV-21, is still a pretty solid robot. Owners generally like it, giving it an average of 4.2 stars on Amazon.com. It earned a “very good” rating at CNET, where reviewer Ry Crist says that it’s “a very effective floor cleaner, earning impressive scores in all of our tests.” (Since our review, Neato has released a new product line called the BotVac, but we still would not choose any of them over the Roomba 650. Most of our problems with the XV-21 will also apply to the new models, with some exceptions.)
The laser-guided navigation system makes the Neato run in a linear back-and-forth pattern, prevents it from running into obstacles or getting stuck in tight quarters, and usually gets the bot to return back to its dock at the end of every session—it got back to its dock 8 out of 10 times over a few weeks of testing. It seems smarter than Roomba’s bump’n’run approach.
On its first few runs, the Neato picked up as much hair and debris as the Roomba 650, but diminishing returns started to kick in after about a week—for two reasons, I believe.
First, it only makes one pass, maybe two, over any spot on the floor during a cleaning session, whereas the Roomba makes three or four (or sometimes none, but it works out better for the Roomba over multiple cycles). Second, the Neato’s nav system is more hesitant to let the bot get into corners or under furniture, which is where debris tends to pile up. It also has no spinning side brush to kick debris away from walls. So, sure, it’ll usually cover most of the visible flooring in your house in a single run and get back to its base. But it’s still leaving a bunch of stuff behind in the hardest-to-reach places.
The interface on the XV-21 is slightly more modern than the Roomba, in that the monochromatic LCD looks like it was recycled from a clickwheel iPod rather than a Memorex clock radio. But functionally, the scheduling and cleaning options aren’t any easier or harder to use than the Roomba. Rather than battery-powered “virtual walls,” the XV-21 comes with magnetic strips that the Neato will avoid crossing—it’s a more elegant solution, and one that doesn’t require stupid D batteries.
In terms of maintenance, the Neato doesn’t need quite as much dirty-hands upkeep—hair doesn’t jam up its moving parts as readily as the Roomba’s, so you really just need to empty the bin, clean the filter, and cut away the odd tangle from the roller bar with scissors. But if a major component breaks down, the Neato isn’t as easy to fix. It isn’t designed to be opened up as easily as the Roomba—there are a handful of screws to get out and a few tiny parts that can fall out and roll away. It’s also harder to pop the rollers in and out, since you’ll need to work around the drive belt (the Roomba uses a belt-less direct-drive system for its rollers).
The Neato also has a tougher time driving up onto area rugs than the Roomba—during every session, the XV-21 bunched up at least one rug, whereas the 650 only struggled with one specific rug. The XV-21 doesn’t have as much clearance to get under furniture as the 650, and some users have complained that Neato bots aren’t aware of their own height (and in situations where they get wedged, the users say the bots can’t comprehend how to steer out of the situation).
A short battery life—either per cleaning session, or the lifespan of the battery—is a common complaint with Neato bots. I found that the XV-21 usually cleaned for about 25 minutes at a time before returning to its station, charging for 45 minutes, and then returning to the spot where it left off. Other owners complained of the batteries no longer holding a charge after 8 to 12 months (and replacements are $80, which is steep). Based on this customer service email shared on RobotReviews.com, it looks like Neato recognized that the 8-month lifespan was a problem, and updated the firmware on their bots to favor shorter cleaning times (25 to 40 minutes is normal) per session so that the battery functions properly over more months. Since the bot does a good job of charging itself and picking up where it left off in the middle of a session, I don’t see a problem with that.
Lastly, the Neato just doesn’t give off the warm-fuzzies that the Roomba does. It’s not there to blend in with its surroundings and become a part of your home. It’s a machine that’s there to do a job, period. Noise is the chief offender—vacuums are loud, but the the XV-21 is deafening for such a small machine, especially compared to the Roomba.
There’s one compelling case for picking a Neato over a Roomba. If your floors are carpeted wall-to-wall and there isn’t too much furniture to steer around, the Neato might be the way to go. It’s still loud and slow, and you’re still going to have problems with the battery. But the vacuum is most effective when its on carpet, and when it can drive in straight lines.
How we tested
CNET provides great data about several bots’ raw cleaning power. The goal of our testing was to see how those numbers translated into the real world, and to get a general idea of what it’s like to live with these things.
I also wanted to get a feel for maintenance: how often do these things need to be opened up and cleaned to work properly, and how much work goes into the upkeep?
I’ve covered most of the specifics above, but I learned a few general lessons that apply to most (if not all) current bots. For starters, they don’t “learn” your house. They just go and run their nav systems every single time you send them out for a cleaning, so there’s no disadvantage to just picking them up and moving them to a different starting point.
Also, as nice as it is to think that the bots should just work on their own, you might have to prep your home before cleanings. Both bots had trouble with one of the light area rugs in my apartment, so I just got in the habit of folding them up and putting them away while the bot ran. Same thing with cords; the robots tangled on charger cables pretty easily (especially the Roomba), so I got used to stowing them away.
As for the dozens of robot vacuums that didn’t make our semifinals…
Miele, maker of the best high-end vacuums, released a robot of their own in 2014. Rather than lasers or bumpers, the RX1 Scout navigates using a camera system. User ratings are generally favorable, but some reviewers wrote that any Roomba they’ve owned was a better cleaner than the Miele. Other commenters have speculated that the scout is just a re-badged Yujin iClebo Arte, which got a mediocre score at CNET.
Neato has a product line called BotVac, with four new models. The BotVac 80 is the best one, because it’s the cheaper of the two models that use a two-brush agitator system. That said, we’d still recommend the Roomba 650 over any of these updated Neato bots. We checked out CNET’s review of the BotVac 85, some early Amazon user reviews, and a deeply informative thread at the excellent Robot Reviews forums Based on all that we’ve read, not enough has changed to fix most of the shortcomings of the XV series.
The navigation system in the BotVac is supposed to be smarter, but the early adopters at Robot Reviews can’t spot the difference. The dimensions are almost identical, so the same obstacles will block it the same way as before. It’s just as loud as its predecessors, and still gets caught on the edges of some rugs. And even though Neato (finally) added a sweeping side brush to the BotVacs, it still doesn’t clean edges as well as Roomba bots do.
And then there’s the issue of price. One of the most popular pro-Neato arguments was that even the best model was cheaper than the low-end Roomba. That’s no longer the case, as the mid-range BotVac 80 now costs $500. Prices are creeping upward without major improvements, the same way they have with Roomba models—it’s not a great trend for anybody.
There are some incremental upgrades in the BotVacs. CNET’s testing found that they clean bare floors better than the XV models did, probably because the brushes are wider and can collect debris from a larger area under the bot. The dustbin is bigger, the filter allows more airflow, and there are some mechanical tweaks that should prevent debris from building up around gears and bearings. And ah yes, the display has leaped ahead a few years—it’s now like the color LCD from a late-decade iPod nano.
We still think that the Roomba 650 is the better pick for almost everyone. And if we were to recommend a Neato, we’d still go with the XV-21, because it’s cheaper.
The Neato XV-11 and XV-12 are the oldest Neato models, and use an agitator design that (as far as we can tell, though there is no independent hard data on this) doesn’t clean as well. They’re also more expensive than the XV-21 at the moment, so no dice.
The XV Signature uses the same agitator as the XV-11/12 and costs more than either, so that’s also a no-go.
The XV Signature Pro is very similar to the XV-21. It ditches the Super Nintendo motif in exchange for a black finish, and comes with two “high performance” filters rather than a single standard filter. It also costs an extra $100 (at the time of writing), which is too much to pay for a boring paint job and a spare part. Both models use the same motor and agitator, so cleaning performance should be nearly identical. CNET found that the Sig Pro, which is a newer model than the XV-21, has a more confident navigation system—even controlling for firmware. If anything, the XV-21’s less-polished mapping tech helps it clean more, since it’s more likely to get a little bit lost, then re-tread ground that it’s already covered. We did some non-scientific side-by-side testing of our own, and did notice a difference, but it’s nothing that changes our main conclusion, which is that you should buy a Roomba.
Neato recently announced the XV Essential, which is yet another iteration on the XV-21 and Signature Pro models. It costs more than the XV-21, and will only be available at Walmart. Pass.
The Roomba 630 is basically the same bot as the 650, but has no scheduling feature, which is a dealbreaker for us.
I strongly considered including the Roomba 760 as a semifinalist, but decided against it. It’s one step up from the 650, only costs an extra $50, and has some features that looked like they might’ve been useful—particularly the Persistent Pass, which will allegedly spend more time cleaning areas that are extra-dirty. I couldn’t find any solid evidence that this is actually useful, though. My hunch is that, since bots aren’t especially great cleaners anyway, it’s like spending time trying to break ice with a snow shovel—it’ll get the job done eventually, but you should just use the right tool (an icebreaker) instead. It also has a remote control, which could be useful, and an extra virtual wall. The dust bin can hold slightly more dirt, too, because it uses a smaller filter than the 600 series. It’s not a bad bot at all, but I don’t think there’s anything here that really makes it a better buy. (The previous Sweethome guide to robot vacs suggested that it might be better for pet owners because of the HEPA filter, but that just didn’t turn out to be true.)
The rest of the 700 series (770, 780, and the now-discontinued 790) are what Cangeloso calls “luxury” robots. Each successive model adds one or two extra features, like a full-bin indicator for the 770, and touch-sensitive controls for the 780, that do nothing to change the cleaning ability. Each tier costs $100 more than the one before it, which isn’t worth it.
The new Roomba 880 is a better vacuum than any of the older Roombas. According to CNET’s testing, the new brush design brings its carpet-cleaning performance up to par with the XV Signature Pro without losing the floor-cleaning prowess of other Roomba models. That said, it’s $700, which is way too much money for something that still is nowhere near as effective as a $200 upright vacuum. As Cangeloso put it in his review of the 880 for Geek.com, “What’s interesting about the Roomba 800 series is that while it’s a much better vacuum than the previous models, it’s not a better robot.” If iRobot puts this cleaning head in its cheaper models down the line, they’ll be onto something.
Moneual makes a few hybrid bots that both mop and vacuum, the Rydis H67 and Rydis H68 Pro. They aren’t very good. CNET absolutely trashed the H67, rating it “Poor” with just 1.5 out of 5 stars. It constantly jammed itself on pet hair and had trouble navigating itself back onto its dock from just a few inches away. The Wall Street Journal doesn’t like the H68 Pro, either. Their testing procedures were sub-par, but even so, the Moneual failed to match the performance of either the competing Roomba or Neato models.
As for other brands, LG and Samsung both have their toes in the water, but their bots are too pricey. It’s $800 for LG’s current Hom-bot, and Samsung doesn’t even list their bots on the US version of their website (imports are about $1,200). CNET reviewed LG’s latest, the Hom-bot Square, but said that “after comparing its price and performance with the competition, it’s a tough purchase to justify.”
The forum moderators from RobotReviews.com that I chatted with spoke highly of the Kärcher RC3000, available only as an import from Germany with a hefty price tag. “The Karcher robot is renowned for being reliable but costs over $1,000. It seems to clean well but that price point is prohibitive,” says Mike Fortuna, one of the forum mods.
CNET reviewed the top-end CleanMate QQ5 from Infinuvo, saying that it “failed to deliver acceptable performance” and adding that “between its hard-to-clean rollers, thoughtless dust bin design, and, in certain circumstances, its inability to clean, we wouldn’t recommend this vacuum at any price.”
CNET also reviewed the newer Infinuvo, the Hovo 510 model. It’s a huge improvement over the QQ5, and keeps pace with Roomba and Neato in most cleaning situations. But it just cannot handle any amount of pet hair—in CNET’s test, no measurable amount made it past the brush roller into the dust bin. If Infinuvo can fix that fatal flaw next time around, it could be a contender—and for much less than any Roomba or Neato model.
What to look forward to
After 14 years of not-so-secret development, Dyson is set to release the 360 Eye robot vac. The marquee feature is a nav system powered by a 360-degree panoramic camera and infrared sensors. It also uses the same powerful motor as some of its older stick and hand vacs. No word on pricing yet, nor a specific date for availability in North America, though it will be sometime in 2015.
At a glance, the vacuum components look great. The brush roller extends the entire width of the bot, which should help it clean up against walls and corners more effectively than a typical bot with a side brush. The V2 motor, also used in the DC34 and DC44 vacuums, is quite powerful.
The nav system looks promising, but this is Dyson’s first bot, so we’re skeptical. Yes, it’s the first bot to use a 360-degree camera. But plenty of others use optical systems—with mixed results. Optical bots work great with open floor plans, but stumble when they have to steer around a bunch of furniture. The Dyson bot is also pretty tall, which means that it won’t get under your couch, and might struggle near any cabinets that jut out a few inches. What’s more, the Dyson’s battery life is only 20 minutes (Roomba is more like 90, by comparison), so the bot will have to be as quick and agile as a human with a DC44 stick vac to clean a space in one pass. Maybe Dyson nailed it, but we’ll need to see it in action—and in any case, we’re pretty confident this will cost a lot more than the entry-level Roomba that we recommend, so there’s no sense in waiting around for this thing.
Wrapping it up
After a few weeks of living with these cleaning bots bouncing around my apartment, I can say for sure that they have made it cleaner, mostly by just doing a diligent job of keeping cat hair off the floor, and finding crumbs brushed away in places I’d never bother to hit with a manual vacuum. They actually do save time and are pretty fun to have around. And they sure as hell beat vacuuming by hand. If you ask us, the iRobot Roomba 650 is the one to buy due to its superior ability to pick up dirt, its more thorough cleaning patterns, its longer battery life when cleaning, and its modular nature that makes the device easy to fix when things go wrong. If you have lighter cleaning needs and don’t have carpets or pets, the iRobot Braava 320 is a more affordable option.
Forum Moderator, RobotReviews.com,
iRobot Roomba 790 review: A charming, low-maintenance little cleaning luxury, CNET, September 8, 2013,
iRobot Roomba 880 review: This bot leaves the competition in the dust, CNET, November 11, 2013,
Neato XV Signature Pro review: Is this boring little bot better than the Roomba?, CNET, September 8, 2013,
Originally published: February 5, 2014