Looking for a robot vacuum? We've bought and performed hands-on tests on over 25 models in the last 5 years. In this 2020 update, we test the best 8 you can buy today. We ran each vacuum through a gauntlet of tests to see how effectively they cleaned the entire room, navigated obstacles, picked up debris (pet hair included). Almost all these models are internet-connected and we assess each smart feature as well. Read on as we describe which vacuums clean effectively and are hassle-free and which ones involve constant babysitting with snags and jams.
The Best Robot Vacuums
$899.99 at Amazon
$799.99 at Amazon
$602.49 at Amazon
$452.99 at Amazon
$299.99 at Amazon
|Pros||Great at picking up pet hair, cleaning carpets, doesn’t miss spots||Great home coverage, excellent navigation, automatically empties its dust bin||Fantastic at cleaning hard and soft floors, packed with smart features||Amazing at navigating, covers all of your home, solid smart connect ability||Great value, excellent cleaning abilities|
|Cons||Exceptionally expensive||Expensive, struggles with pet hair||Pricey, struggles at navigating in confined places, little rough on furniture||So-so carpet cleaning skills, below average at cleaning hard floors, poor at picking up pet hair||Good, not great at navigating|
|Bottom Line||If you are looking for a premium robot vacuum that can clean up after your fur-babies, the Roomba s9 is a great bet||If you want a top-notch robot vacuum that even empties itself, then the i7+ is a great choice, but expect to pay for it||If you have heavy messes to clean up and want a robot to do it for you, then the D7 is the perfect choice||This robot does a phenomenal job at navigating through even the most cluttered homes, but definitely could do a bit better when it comes to cleaning performance||If you want the most bang for the buck when it comes to these products, then this is your best bet|
|Rating Categories||iRobot Roomba s9||Roomba i7+ with...||Neato Botvac D7...||Xiaomi Mijia...||Neato Botvac D4...|
|Home Coverage (20%)|
|Carpet Cleaning (15%)|
|Hard Surface Cleaning (15%)|
|Pet Hair (10%)|
|Smart Connect (10%)|
|Specs||iRobot Roomba s9||Roomba i7+ with...||Neato Botvac D7...||Xiaomi Mijia...||Neato Botvac D4...|
Clean base holds 30 bins worth
|Dimensions||12.4" x 12.2" x 3.6"||13.34" diameter
Charging base: 12.2" W x 15.1" D x 19" H
|12.5" x 13.2" x 3.9"||13.9" diameter
|12.5" x 13.2" x 3.9"|
|Measured Weight||9.2 lbs||7.4 lbs||7.9 lbs||7.8 lbs||7.4 lbs|
|Warranty||1 year limited||1 year limited||Robot - 1 year limited
Battery - 6 month limited
|1 year limited||Robot - 1 year limited
Battery - 6 month limited
|Runtime (manufacturer claimed)||Up to 120 minutes||Up to 75 minutes||Up to 120 minutes||Up to 150 minutes||Up to 75 minutes|
|Recharge & Resume||Yes||Yes||Yes||Yes||Yes|
|Charge time||2-3 hours||2-3 hours||2-3 hours||Up to 4 hours||2-3 hours|
|Multi-Room Navigation (Cleaning)||Yes||Yes||Yes||Yes||Yes|
|Barrier Type||Virtual (not included)||Virtual (comes with 1)
Room by room cleaning (in App)
|No Go Lines (in App) and Physical (magnetic strip included)||Magnetic Strip (Not yet available)||No Go Lines (in App) and Physical (magnetic strip not included)|
|Filter Replacement Cost||3 for $40||3 for $25||Standard 4 for $20
High Pre 2 for $20
Ultra 2 for $30
|Washable, comes with 2||Standard 4 for $20
High Pre 2 for $20
Ultra 2 for $30
|Brush Width||9 5/8"||6 3/4"||10 3/4"||6 1/2"||10 3/4"|
Best of the Best for Most People
iRobot Roomba i7+ with Clean Base Automatic Dirt Disposal
If you are on the hunt for the best of the best when it comes to these products, then it is hard to go wrong with the Roomba i7+ with Clean Base Automatic Dirt Disposal. This robot thoroughly impressed us with its navigational skills, cleaning even the most cluttered areas with relative ease and seldom requiring intervention to free it. It also does a great job of cleaning up all sorts of messes from hard floors or flat or fluffy carpet. This little robot is jam-packed with smart features and functions, allowing you to command the i7+ remotely and integrate with existing smart home systems. On top of all that, the automatic dirt disposal base means you only have to clean out the dustbin after 30 cleaning passes or so, instead of after every single time you run the Roomba, making it even more convenient to use this automated floor cleaner.
However, we did find one noticeable drawback with this robot vacuum. It did an awful job in our pet hair test. This robot barely collected any of the pet hair we spread out on either the low-pile or medium-pile carpet and the vast majority of the hair collected was tangled up in the extractor head, rather than making it into the collection bin. Regardless, this is still one of our favorite robot vacuums and would make a great addition to any pet-free household — or to a house with pets, provided you don't mind a little supplemental vacuuming!
Read Review: Roomba i7+ with Clean Base Automatic Dirt Disposal
Best All-Around Vacuum
Neato Botvac D7 Connected
The newest flagship of the Neato line, the Botvac D7 Connected delivered an overall fantastic performance, especially in our cleaning tests. This smart vacuum will create maps of your home after it cleans, allowing you to track its progress and even draw virtual walls to delineate areas where you don't want the robot to venture into it. It does an excellent job of cleaning up most types of mess from both hard and soft floors an even does quite well at picking up pet hair.
The Neato D7 isn't without its flaws. We found that this robot can be prone to get caught or stuck whenever it is cleaning in more cramped areas around your house. Additionally, it is a little less gentle on your furniture than other robots, especially when backing up. The app does have a "Gentle Navigation" mode if this is a problem, but we didn't notice a huge difference. The D7's top-tier performance also comes at a bit of a price and is one of the more expensive models of the group. It's easily one of our all-time favorite robot vacuums we have tested to date and we would highly recommend it to anyone searching for the best of the best.
Read Review: Neato Botvac D7 Connected
Best Navigational Abilities
Xiaomi Mijia Roborock V2
Matching the overall score of the D7, the Roborock V2 by Xiaomi distinguished itself by being the best navigator by far. This robot easily flitted in and out of the most confined areas around various types of furniture, never once getting stuck or requiring assistance to be freed in our tests. It covers the largest percentage of your home out of any of the other products we tested, never omitting an area if the Xiaomi physically could fit there. It also has an impressive set of smart features.
However, the Xiaomi doesn't come close to matching the cleaning abilities of the Neato Botvac D7, struggling to pick up pet hair and other caked-on messes. While it doesn't top the charts when it comes to cleaning power, this robot is by far the best at making its way through a cluttered and cramped home, boldly cleaning where other robots fear to tread.
Read Review: Xiaomi Mijia Roborock V2
Best Bang for the Buck
Neato Botvac D4 Connected
If the cost of the top-tier robots is way outside your budget, then the Neato Botvac D4 Connected should be the next model you look at. It's our absolute favorite of the robots in its price range, even performing comparably to some of the overall top robots in some of our tests. This vacuum navigates obstacles exceptionally well and does a good job cleaning the majority of your home, picking up all sorts of messes from both carpets and soft floors. It even does a fairly good job collecting dog or cat hair and has a respectable amount of smart features and functions, all while retailing at a fraction of the price of the premium products.
However, the Neato D4 does encounter some difficulties when trying to clean confined or cluttered areas of your house, skipping the area entirely if it detects too many obstacles. The D4 also isn't exactly gentle when driving around your furniture, with it repeatedly backing into items in our test course or jostling them as it tried to find a suitable route. The most this would ever amount to is some light scrapes and scuffs but might be something to consider if you have particularly delicate or antique furniture items in your home. This robot also didn't do quite as well when it came to cleaning performance as the top models, especially on the hard floors. However, the D4 is by far the best you can get if you are looking to save some cash but still want one of the best 'bots around.
Read Review: Neato Botvac D4 Connected
Best on a Tight Budget
Ecovacs Deebot N79S
If the D80 is still outside your budget and you have to have a robot vacuum, then you should consider the Ecovacs Deebot N79S. While this robot left much to be desired, it does a decent job of navigating your home and has a respectable set of smart features. It isn't the best when it comes to cleaning performance but it will clean up small messes without too much of a fuss but its main selling point is that it costs considerably less than the top products.
The N79S has a hard time cleaning up stuck-on messes or particularly difficult types of debris, like pet hair or flour. This robot also can't reliably clean large rooms or multiple rooms at a time. However, if you want a robot vacuum for light messes or cleaning a single room and are shopping on a tight budget, the Deebot N79S is our top recommendation for you.
Read Review: Ecovacs Deebot N79S
Top Pick for Pet Hair
iRobot Roomba s9
While the Roomba s9 did merit the top score overall, it isn't necessarily the best option for most people — even for those that do want the best of the best. This robot is the best of the group that we have seen when it comes to picking up pet hair and did exceptionally well at cleaning up all sorts of different messes from both hard and soft floors. It has a decent set of smart features and will usually cover the vast majority of the accessible floor area when it completes a cleaning cycle. It also navigates quite well but is far from the best. You also can purchase an automated dirt disposal base — similar to the i7+ — so you only have to empty the base every 30 cleaning passes or so. However, this adds a non-trivial amount to an already pricey product.
This robot can get a little tripped up around densely packed furniture on occasion — especially when it is the first time that it is cleaning a room — and can't come close to matching the finesse the Xiaomi moves around obstacles. The s9 did get better over time but we didn't think it navigated quite as well as the i7+ did in our tests. The main reason we don't recommend the s9 to everyone is it borders on being prohibitively expensive, especially since it still isn't good enough to completely supplant a traditional vacuum in our opinion. It's a fantastic option if you can afford it and need to pick up tons of dog hair but we think most people will be happy with some of the other vacuums that scored only a little bit lower and cost considerably less.
Read Review: iRobot Roomba s9
Why You Should Trust Us?
We have spent over three years testing and comparing robot vacuums, ranking and scoring the performance of all the best 'bots and updating our review as new models have been released. Here at TechGearLab, we won't ever accept any free or evaluation units, so you can be fully confident that we are completely unbiased and don't have any incentive financially to pick one product over another. We bought all of the robots in this review at normal retail pricing, just like you would. Our review team is composed of Austin Palmer and David Wise , who both have extensive experience with consumer electronics, having tested and reviewed hundreds of different tech and smart home products.
We spent close to 250 hours testing these robots side-by-side, even going so far as to make simulated furniture in our test room so we could directly compare the home coverage of each robot with long exposure photography. We spread out controlled amounts of messes on different surfaces for our cleaning tests, then scored each robot on how much it picked up. We even got pet hair donated from a local groomer to spread out on our test carpet for our pet hair collection tests. Finally, we also rated the reliability of the network connection by seeing how many times we had to reset it throughout our testing period and compared how easy and intuitive the remote interface of each robot vacuum is.
Related: How We Tested Robot Vacuums
Analysis and Test Results
We divide these tests into six weighted rating metrics — Navigation, Home Coverage, Carpet Cleaning, Hard Surface Cleaning, Pet Hair, and Smart Connect — each weighted based on their overall significance for these products. Additionally, we also consider the price of each product and how that connects to its performance, for those of you shopping on a tighter budget and looking for better value options.
Related: Buying Advice for Robot Vacuums
Right off the bat, it's clear that you are going to have to spend some cash for a top-notch robot vacuum. If you are shopping on a budget, then our top recommendation is the Botvac D4 Connected by Neato. We found this robot vacuum to offer the best value without sacrificing too much in terms of performance, doing a great job at navigating your home and cleaning up most types of messes, all while being a few hundred dollars less than the flagship models. If this product is still too pricey, then the Ecovacs Deebot is your best bet if you are shopping on a tight budget and just have to have one of these products. However, there is a significant drop in performance when comparing this model to the D4 or our other top award winners. If you are searching for the best of the best, you can expect to pay a decent amount, but these premium robots will do a much better job of cleaning and become stuck much less frequently than the less expensive products.
Our most important testing metric, Navigation constitutes 30% of the overall score for each product. The majority of each robot's score for this metric is based on how it performed in our furniture navigation challenge. For this test, we constructed a set of simulated furniture — dining room table, four chairs, a lamp, sofa, coffee table, and a comfy armchair — and set them up in a room to see how well each robot handled cleaning around them. We were primarily focused on if the robot could navigate the room without assistance, rather than on how much of the room it cleaned. Robots also lost points for exceptionally erratic behaviors, such as repeatedly cleaning under a table for the bulk of the time. Our next metric covers the effectiveness of each robot at cleaning your home, so a robot failing to clean under a piece of furniture would be penalized there.
In addition to this navigation test, we also compared and scored the docking abilities of each robot, testing if they could find their way back home from a different room, and how well each one navigated over a dark carpet. Some of the edge detection sensors can read a false-positive when driving over the transition between light and dark carpets, causing the robot to refuse to clean darker carpets and back away from them repeatedly.
Earning the top score overall and delivering by far the best performance that we have seen to date, the Xiaomi Mijia Roborock V2 earned a 10 out of 10 for its essentially flawless performance in these tests. This robot easily navigated around all of our test furniture in each of the three trials that we conducted, failing to get stuck in any of them.
This model drove in and out of the tightest of spaces in our test furniture layout and was also very delicate in its maneuvers. Other robots would be much rougher with the furniture and back into it quite aggressively, where the Xiaomi would very carefully navigate around it and negotiate tight spots without crashing into it.
This robot also handles docking exceptionally well, quickly and easily finding the dock when it is in the same room as the robot. On top of that, it also delivered the best performance we have seen at locating its dock and returning home when the dock is in a different room — a much more difficult test that stymied a handful of robots. It also cleans darker carpets with ease, with the edge sensors failing to falsely trigger a single time in any of our tests.
Following the Xiaomi, the Roomba i7+ claimed the second place position when it comes to navigation, meriting an 8 out of 10. This robot handled our testing room full of furniture with ease, never getting completely stuck or being overly rough with any of the items. It did have to take a bit longer in some of the more cramped areas to locate a way out, while the Xiaomi immediately located an exit route.
However, it also was completely unaffected by areas of high-contrast in our test, cleaning the dark rug on the light carpet without issues and can easily find its way home to its dock, regardless if it was in the same room or a different room when commanded to return.
Following the Roborock and the i7+, the Roomba 960 and the Roomba s9 earned a 7 out of 10 and tied for the third-place position in this group of evaluations. The 960 never got stuck in our furniture test, but navigated the test room with noticeably less finesse than the Roborock — on par with the i7+.
The 960 would occasionally make a bunch of extra passes around various furniture items and would struggle a bit more in finding a clear route out from furniture when they were surrounded, such as under chairs, but always finished its cleaning passes reasonably quickly. We also noticed that it is a bit rougher on the furniture than the Xiaomi Roborock, relying a bit more on the contact sensor than the vision sensor when cleaning around smaller obstacles, such as our fake chair legs.
The Roomba 960 also did very well in our docking tests, beelining straight for the charging base regardless if they were in the same room or a different room. However, it wouldn't drive over the border between light and dark carpeting, thinking it was a drop-off.
The Roomba s9 also didn't have any issues docking, almost always finding its base immediately and it had no issues driving over high-contrast areas of your floor, unlike the 960. The edge detection sensor never gave us a false-positive in the course of our testing, only registering a drop when there was one.
Unfortunately, we found the s9 to be a little finicky when negotiating its way around furniture. Its first run in our test course was abysmal but its performance improved significantly after we ran a training run and started it over. It handled tight areas like a champ but would occasionally crash into obstacles head-on — similar to the Neato robots. It also got tripped up in some of the tighter areas, struggling to find a way out around the chairs in our fake dinette set.
After this top group of vacuums, the bulk of the robots all scored an overall 6 out of 10 for this metric. This group is comprised of both systematic and random navigation robots.
Of this group, the Botvac D7 Connected and the Botvac D4 Connected by Neato are the systematic cleaners. We were a little disappointed with the performance of this duo in our furniture navigation challenge, as they all required some amount of human intervention to complete their cleaning cycles. These all choked up around the simulated dining room table and chairs, becoming entrapped in the tight spots.
Also, these were exceptionally rough with furniture, even pushing some of the items around as they struggled to free themselves.
However, this pair did redeem themselves in the other two tests in this metric, heading straight for the dock when directed to return home and being unaffected by the transition between light and dark floors.
The remainder of the robots in this group — the Ecovacs Deebot N79S and the Shark ION Robot 750 all essentially clean your room randomly. They do avoid and navigate in and out of obstacles but don't map out their overall cleaning plan. Surprisingly, these robots are quite adept at not getting stuck, though it may take them a substantial amount of time to locate a clear path to free themselves from tight spaces. It's essentially trial and error for them to locate the correct path, but none of these robots ever needed our assistance if we were patient enough or they would simply avoid any confined areas and skip cleaning them. All three of these robots also successfully navigated the transition from light to dark flooring without an issue.
However, neither the Ecovacs nor the Shark were particularly adept at navigating back to their docks in our tests. This trio would do alright if they were sent home while they were in the same room as the base, only bouncing around for a bit before finding it and returning home, with the Ecovacs Deebot taking the least amount of time. If any of these robots were in a different room, they would eventually find the base, but it could take quite a bit longer than the other vacuums, on the order of 3-5 minutes.
Finishing at the back of the group, the Samsung POWERbot R7070 earned our lowest score of 5 out of 10 for this metric. This robot suffered quite a bit in our furniture test, avoiding most of the problem areas right from the start. Despite that, it still managed to get stuck to the point of needing assistance to be freed, as well as being trapped in a — seemingly — endless loop the one time it did venture under the table. You can see in the time-lapse photo below that it overexposed the shot under the table with its repeated passes as it struggled to free itself.
It also refused to cross the transition between light and dark carpet, fearing that it was the edge of a step or other drop off, but it does navigate back very quickly to its charging base when commanded to.
Following our Navigation metric, Home Coverage is the next most important set of tests, accounting for 20% of the overall score for each vacuum. For this metric, we evaluated and scored how much of your home each robot could clean, rather than just seeing how well they could navigate autonomously. We tested the size and ease of use of the spot cleaning functions of each robot, as well as their abilities at cleaning multiple rooms and recharging if necessary. We also looked at how much of the room each robot cleaned in our furniture test, as well as the effectiveness and ease of use of barriers to create a no-go zone and cordon off areas that you don't want the robot to clean. This can be particularly useful in areas where the robot is likely to become trapped or suck up something it shouldn't, such as around a shoe rack or pet food bowls.
The Xiaomi continued its reign in this metric, again earning the top score of 8 out of 10, but was also joined at the top by the Roomba i7+ and the Roomba s9. The Xiaomi repeatedly cleaned every last spot around the furniture in our test, never avoiding a location because it was too cluttered and only failing to clean areas that were physically impossible to fit the robot. This robot also handles multi-room cleaning very well, finding its way easily between rooms in a systematic way. The Xiaomi will also stop cleaning, return to its base to charge, and resume cleaning automatically.
Unfortunately, the Roborock doesn't include any barriers, requiring you to purchase them separately. It utilized magnetic strips as a way to signify no-go zones. This isn't our favorite method, as it is hard to place them securely and discreetly, but they are effective. Finally, we evaluated the spot clean function of this product. It's only mediocre when activating it from the robot directly, doing a single pass over a square area about 5.5 sq. ft. in size, but it is excellent when using the companion app on a mobile phone.
The app allows you to set up to five zones in any of the rooms that the robot has mapped of any size and initiate a spot clean of them. The robot will travel to each of the zones and clean them with the prescribed number of passes, ranging from 1-3. This is particularly handy, as we found that the Xiaomi didn't quite pick up as much debris as the other products did in a single pass, making it quite handy to set it to complete multiple passes.
Both the i7+ and the s9 performed almost identically in our home coverage tests, even though we thought the i7+ had a slight edge in our previous navigation tests. This pair does a fantastic job of cleaning multiple rooms, systematically moving from one room to the next and creating a map of your home while doing so. Both 'bots monitor their battery levels while cleaning and will pause their cleaning cycle to dock and recharge if they are going to run out of juice. They will then automatically resume cleaning where they left off once the battery is recharged. The i7+ will even empty its dirt disposal bin into the automatic base, saving you from having to empty it daily.
Both the s9 and the i7+ will do their best to clean any confined area that they will physically fit into — though the s9 never exhibited the same finesse the i7+ did when leaving these areas. The i7+ included a single virtual barrier system if there is an area you don't want it to clean — such as continually crashing into your shoes left by the door or a pet's food and water bowls. The s9 didn't include any sort of barrier system but is compatible with the same one as the i7+; you just have to purchase it separately. Additionally, both of these robots let you create virtual barriers and no-go zones in the app once they have mapped your home, which is quite a bit more convenient than a physical barrier or virtual wall with a base station.
We were very impressed with the spot cleaning function on both the i7+ and the s9, which is comparable to the Xiaomi Roborock. While the spot clean initiated using the button on either Roomba is fairly mediocre, sending the robot to clean a circular area about 4' across, the spot clean initiated from the mobile app is much more capable. You can send these robots to clean specific rooms once they have created a map of your home.
A group of products tied for the runner-up position, with the Neato Botvac D7, D4, and the Roomba 960 all meriting a 7 out of 10. All of these robots are highly capable when it comes to cleaning multiple rooms of your home, systematically cleaning each room before moving on to the next one. These robots will also all return to their docking station to charge if their battery becomes depleted while in the middle of the cleaning cycle, automatically resuming once they have recharged sufficiently.
The Roomba 960 covered slightly more ground than the Neato robots did in our single-room furniture test, nimbly cleaning around all of our simulated furniture.
The Neato D7 and D4 tend to struggle a bit more in confined areas, getting jammed up and failing to clean them. The 960 would happily venture where the Neatos would not.
However, we did like the barrier methods on the Neato D7 and D4 the best, followed by the Roomba 960. Both Neato robots rely on a physical, magnetic strip to signify no-go zones but also allows you the ability to set virtual barriers in the mobile app.
This can take a little bit of time to set up, requiring you to run a special cleaning cycle and follow a series of prompts, but they are quite useful in practice.
The Roomba 960 has a virtual barrier system that can be placed on the ground. This small box can either generate a virtual wall about 10' long or a do not enter circle about 4' in diameter, identical to the i7+.
For our spot clean check, we found the Roomba 960 to have a far inferior spot clean compared to the Neato D7 and D4, only cleaning about 12 sq. ft. This robot would spiral out and in from the location where you initiated the spot clean and didn't do an amazing job at collecting debris.
The Botvac D7 and D4 both normally clean a square section that measures roughly 49 sq. ft. with their normal spot cleaning modes. However, both of these robots can be set to clean a much larger area in their companion mobile apps, boosting the coverage to almost 170 sq. ft. — though you still can't set them to perform multiple passes.
The Samsung POWERbot R7070 came next, delivering the worst performance out of any of the systematic cleaning robots in this metric and earning a 6 out of 10 for its efforts. This robot seems to very purposefully avoid areas that are cluttered and can become quite easily trapped if it does venture into them, resulting in many areas of your home not being cleaned while others may be cleaned relentlessly. This robot also utilizes magnetic strips as barriers, without any option to set up virtual ones in the app. Additionally, we also weren't enamored with the spot clean feature on this product. While it did clean a comparable area to the Botvac D4, it left a decent amount of debris behind in our tests.
However, this robot does handle cleaning multiple — relatively clutter-free — rooms quite well and does have the ability to pause cleaning to recharge and automatically resume, like the other top products.
Finishing out the back of the group for this metric, the Roomba 690, Shark ION, and Deebot all earned a 4 out of 10. All of these products clean in a random fashion, meaning that none of them particularly impressed us when it came to cleaning our furnished room. These robots all bounced around, missing quite a few spots unless left to clean for an exceptionally long time — orders of magnitudes more than the systematic cleaning robots.
The Shark ION uses a magnetic strip to define no-go zones — again, not our favorite method — while the Roomba 690 uses the much more preferable virtual barrier methods. The 690 uses the same item as the 960 or the 980, allowing you to create a virtual wall or circle. The Deebot lacks any sort of way to specify no-clean zones.
None of these robots can stop cleaning to recharge and resume and all have mediocre spot cleans of about 12-15 sq. ft.
After all of our tests assessing how well each robot moved throughout your home, we moved on to scoring how proficient each robot is at cleaning your floors, starting with carpet. We tested on both low-pile and medium-pile carpets, using flour, rice, oats, and Mini-wheats as our sample messes. Additionally, we also scored how closely each robot could clean against the walls and edges of a room. This metric is responsible for 15% of the overall score for each vacuum.
Toppling the Xiaomi, the Botvac D7 and the Roomba s9 both tied for the top spot when it came to cleaning soft floors, both, earning an 8 out of 10 for their superb performance at keeping carpets clean. The Botvac D7 started with an excellent showing in our edge cleaning tests, cleaning within an inch of the walls, but leaving a bit more leftover mess in the corners, leaving a small wedge of leftover debris about 1.5" from the wall at its widest point.
Next, we tested out how well the D7 did at sucking up flour. We only did this test on the low-pile carpet, as expecting any of these products to collect flour from fluffier carpet is a tall order — a task much more suitable for a more powerful stick or upright vacuum. This robot picked up the majority of the flour, showing a clear distinction between the areas it cleaned and the ones it didn't.
The D7 delivered a phenomenal performance in collecting rice on both flat and fluffier carpet, collecting the vast majority of the grains — leaving less than 10 rice grains behind.
The results were the same when it came to picking up oats, with the D7 delivering a top-notch performance, easily collecting all of the mess.
For the final test of this metric, mini-wheat collection, the D7 faltered slightly. It usually got most of the mini-wheats, but there was always at least one becoming jammed or just getting pushed around by the vacuums.
The s9 did even better than the Neato D7 at cleaning in close to the edges of our robot test pen, only leaving a border of uncollected coffee grounds that measured about 0.5" across. The side brush doesn't do all that much at getting into the carpet piles and it takes a little bit of time to right itself coming out of a corner but it overall does a very impressive job at cleaning close to walls.
The s9 also bested the D7 at cleaning flour from low-pile carpet, leaving only trace amounts behind, but the D7 did a bit better at collecting the rice and oats. The s9 matched the D7 at getting the rice from the flatter carpet but left quite a bit more behind on the fluffier carpet. The s9 also didn't pick up as many oats from the low-pile carpet as the D7, leaving a handful of whole oats and a decent amount of crumbs behind but the s9 did get just a few more oats than the D7 did on the fluffier carpet. The s9 also did about the same or just a tiny bit better at sucking up the Mini-Wheats.
Following the top performers, the iRobot Roomba i7+ and the Botvac D4 Connected came next in terms of carpet cleaning prowess, both earning 7 out of 10 for their great showings.
The D4 started with a fantastic performance in our edge cleaning test, matching that of the D7. The Roomba i7+ struggled a little bit more, leaving a strip of residual debris about 3.5" wide around the edges of our test pen.
The round design also meant that it left quite a bit more area uncleaned in the corner, measuring about 5" across at the widest point. Neither of these robots could match the performance of the D7 when it came to cleaning up a floury mess, with the D4 leaving a noticeable amount of flour behind and the i7+ leaving significantly more.
Both the D4 and the i7+ redeemed themselves in the rice and oat cleanup test, delivering excellent performances on both flat and fluffy carpet — although the i7+ did leave a tiny bit more residual rice behind on the fluffier carpet than we would have liked, about ⅛ of a teaspoon.
These both finished out this metric with an above-average showing at sucking up mini-wheats, with the D4 tending to miss a few and the i7+ usually only leaving a fragment of one behind.
Next, the Roomba 960, Roomba 690, and the Samsung POWERbot R7070 all earned a 6 out of 10 for their solid cleaning performance. Starting with our edge test, both the 960 and the 690 didn't do an amazing job, matching the performance of the 980. However, the POWERbot is by far the best out of any vacuum that we have seen at cleaning the corners of a room. This robot left a strip of leftover debris less than an inch wide and cleans right into the corner, leaving essentially no residual mess.
The Samsung delivered another solid performance in our flour cleaning assessment, outperforming the 690 and the 960, but falling short of the Neato D7.
The 690, 960, and the Samsung all did quite well in our rice test, collecting almost all of the debris on the flat carpet and only leaving a tiny bit of rice behind on the fluffier carpet.
The 690 and the 960 did equally well at collecting oats, while the Samsung fell flat in this test, doing very poorly on both the low-pile and medium-pile carpet.
This robot pretty much did the worst of the entire group, leaving behind tons and tons of oats. The 690 and the 960 finished out this metric with an excellent performance in our mini-wheat collection test, easily collecting all of them without issue. The POWERbot didn't fare quite as well, usually leaving 1 or 2 behind, regardless if it was on the flat or fluffy carpet.
The Shark ION Robot and the Xiaomi came next, both earning a 5 out of 10 for their rather middling performance. The Shark did quite poorly at cleaning edges and corners, leaving a considerable amount of debris behind. The Xiaomi did a little better, but not by much — quite a fall from grace for this robot after its unmatched performance in the prior two metrics.
The Xiaomi and the Shark again disappointed in our flour collection test, both leaving behind quite a bit of residual flour, the Shark slightly more so than the Xiaomi.
These robots redeemed themselves in our rice collection test, doing a great job at getting the bulk of the rice and an average one in our oat collection test, leaving a bit more debris behind on both flat and fluffier carpets. These finished out with a substandard performance in our large object collection test, being very hit or miss when it came to sucking up the mini-wheats.
Finishing out the back of the pack, the Ecovacs was the lowest score of the group, earning a 4 out of 10 for their sub-par carpet cleaning abilities. These both didn't clean in very close to the edges of the room, leaving a strip of leftover mess that was at least 3.5" wide. It also left a large wedge in mess in the corners of our testing pen, measuring over 6" at the widest point.
This pair also delivered a below average in the majority of our other tests — oat, mini-wheats, and flour — only doing a decent job at collecting rice. Both of these products retail at a much lower price than the top products but it does show when compared to the top products.
Hard Surface Cleaning
Similar to our set of soft floor cleaning tests, our Hard Surface Cleaning metric is also responsible for 15% of the total score for each robot vacuum. We repeated the same tests as above, though this time using a section of hardwood laminate floor. Cleaning hard floors is much easier for these products, meaning that many of these products scored quite a bit better.
The D7 and the Samsung POWERbot both earned an 8 out of 10 for their stellar performances. These robots all did about the same in our edge test on hard floors as they did in the carpeted version, with the Neato robot leaving a little bit in each corner and the Samsung doing the best by far of the entire group, getting practically everything.
The POWERbot continued its dominance in our flour collection test, collecting virtually all of the flour we laid out and outperforming all of the vacuums in the entire group. The D7 did a little better than average, leaving a little bit of flour on the surface and failing to collect any of the flour that had fallen in the cracks between the boards.
This duo both did an excellent job at collecting rice and oats, though the Samsung did leave behind 3 leftover oats. However, the R7070 did an abysmal job at collecting mini-wheats, pushing them around indefinitely, rather than collecting them. The D7 fared quite a bit better, though still not great, usually only collecting slightly less than half of the mini-wheats we placed on the floor.
Following the first place pair, both the iRobot Roomba s9 and the Roomba i7+ earned a 7 out of 10 for their almost top-tier results. The s9 did much better than the i7+ and even a bit better than the Neato D7 at cleaning along the edges of a room but couldn't quite compare to the Samsung R7070, leaving a strip of flour about a half-inch across behind after it cleaned.
The s9 also did a phenomenal job at cleaning up the flour from the hard floor, only leaving the finest residue behind, which was a bit better than the i7+. Unfortunately, the s9 only did a so-so job at sucking up the oats and dried rice. The s9 picks up the bulk of the mess without any issue but the brush rollers can throw some debris around when it hits a large pile of rice or oats. This usually isn't too much of a problem, as the rice or oats tend to get thrown into the area the robot is about to clean but it does periodically throw some to the side and fail to collect them.
The s9 did manage to collect ⅔ of the Mini-Wheats we spread out. It has plenty of clearance to roll over the cereal but wouldn't always suck it up into the collection bin when it passed over.
The i7+ delivered one of the best performance out of the round robots at cleaning close to the edge in our robot pen, but still left a strip of residual mess up to about 4" wide. It also did a very good job of cleaning up flour, getting most of it on the surface, but failing to really clean into the cracks between floorboards and tracking a little bit of flour around with its wheels.
The i7+ did an awesome job in the next two tests, collecting all of the rice and oats with ease — outperforming the s9. Regrettably, it did fail to collect any mini-wheats, earning a poor score for that final test.
The Neato D4 Connected followed, receiving a 6 out of 10 for its solid carpet cleaning abilities. It got off to a strong start in our edge cleaning test but didn't do quite as well as the D7, as it lacks a rotating side brush. It also didn't do well at collecting flour, leaving behind a visible residue across the surface and failing to get any of the flour that fell between the cracks. However, its performance did rebound when tasked with collecting rice or oats, only leaving trace amounts behind. Unfortunately, it also struggled with mini-wheats, failing to collect a single one.
Next, the Shark ION Robot and the Robot 960 both earned a 5 out of 10 for their average showing at cleaning hard floors. These both left reasonably wide strips of residual debris in our edge test, with the Roomba 960 leaving a considerable amount of brush marks behind.
The Roomba 960 also did very, very poorly in our flour test, leaving behind tons of flour. However, at least it didn't drag the flour around that much, simply leaving it be. The Shark did a little better but failed to clean anything besides the surface flour.
The 960 did redeem itself in our rice and oat collection assessments, delivering an excellent performance and collecting all of the debris that we laid out. The Shark did very well in the oat test but did quite poorly at collecting rice, leaving over an eighth of a teaspoon behind.
Both of these vacuums did very poorly at collecting larger pieces of debris, pushing around all of the mini-wheats in our test indefinitely, as it lacked the clearance to successfully suck them up.
Finishing last in the group, the Ecovacs Deebot, the Roomba 690, and the Xiaomi Roborock all earned a 4 out of 10 for their rather undesirable performance when it came to cleaning hard floors. The Xiaomi did the best of this group in our edge cleaning test, doing a fairly good job overall and leaving a strip of leftover mess about 3.5"-4" across.
The Ecovacs left slightly wider strips of leftover mess. None of these vacuums did very well in our flour cleaning test, with the 690 doing the best and the Xiaomi doing the worst. However, they were all less than average overall.
The Xiaomi, as well as the other three robots in this group, all improved significantly in our rice collection tests. The Xiaomi collected almost all the rice, only leaving a few grains behind. The Roomba 690 did almost as well, though it lost some points due to its rotating side brush tending to fling the rice around. The Ecovacs flung around a bit more rice than the others.
For our final evaluation of the cleaning performance of each of these products, we ranked and scored how well each one cleaned up pet hair. This metric is responsible for 10% of the score, with some robots handling our tests far better than others.
To test these products, we used pet hair kindly donated by a local groomer, then spread a measured amount out on both low-pile and medium-pile carpet. Each vacuum was scored on the percentage of hair that was collected, as well as how much of the hair ended up in the collection bin, rather than tangled up in the brush or on the undercarriage of the robot. This is a somewhat hard test for these robots and is a task much more suited for a more powerful upright or stick vacuum, so the numbers are a little on the lower side, but you could always run the robot vacuum multiple times if you must use them to clean up pet hair.
Easily wiping the floor with the other robots, the Roomba s9 is the best robot vacuum we have seen so far when it comes to picking up pet hair, earning the top score of 9 out of 10. This vacuum collected a whopping 53% of the hair we spread out on the medium-pile carpet and 50% of the hair from the low-pile carpet. Even better, essentially all of this hair ended up in the collection bin of the robot, with little to none getting tangled around the extractor head.
The Samsung POWERbot, Neato D7, and the Neato D4 all tied for the second place position, each meriting a 6 out of 10 for their decent attempt at picking up pet hair. These three robot vacuums collected around 25-35% of the hair we spread out on the low-pile carpet and between 45-65% of the hair from the medium-pile carpet.
The remaining robot vacuums scored very poorly, with the i7+, the Ecovacs, the Roomba 960, the Roomba 690, the Shark ION, and the Xiaomi all meriting a 2 out of 10. These all collected less than half of the hair that the top vacuums collected.
Our last set of tests for these vacuums focused on the various smart features and functions that are prevalent in the vast majority of these products, which accounts for the remaining 10% of the total score. In particular, we looked at how each vacuum can interact with different smart home systems, how intuitive and easy to use the companion smartphone app is, and its functionality, as well as the reliability of the network connection throughout our tests.
The Xiaomi, the Neato D7, the D4, the Roomba s9, the i7+, and the Roomba 960 all tied for the top score of 7 out of 10. All three of the iRobot Roomba robots performed very well in our WiFi reliability test, as we never had to reset the network or restart the app to establish a connection throughout our testing period. The Xiaomi did a little worse, as we occasionally had to quit the mobile app and restart it before the robot would appear as connected. We found the Neato D7 to be very problematic, requiring multiple resets and we struggled to get it to reliably connect, though we didn't suffer any network issues with the D4.
However, the Neato D7 and the D4 did have the most functional app of the entire group. You can adjust the suction power, view a cleaning map, set maintenance reminders, control the robot and draw virtual barriers. This is a unique feature to these products and is exceptionally convenient, giving them a slight edge over the Xiaomi and the Roomba apps.
The mobile apps for the other three robots are quite functional. You can manually drive the Xiaomi around, adjust the suction levels, and view a cleaning map of your home.
The companion mobile app for the Roomba robots also allows you to view a cleaning map of your home, as well as track the life of various accessories on the robot, so you know when you should replace them. Additionally, you can also command the i7+ and the s9 to clean specific rooms, once your home has been mapped.
It's very easy to set a schedule on all of these robots on the app, though you can't set it on the robots themselves. However, you can send the robot home or start a spot clean on the robot itself. The D7, the D4, the Roomba s9, the 960, and the i7+ are a little more fully-featured when it comes to interacting smart home ecosystems, working with Alexa, Google Home, and IFTTT, while the Xiaomi only worked with Alexa, but we found it to be a bit flawed in practice and quite prone to errors.
Following this top group of products, the Ecovacs Deebot, the Roomba 690, and the Shark ION all earned a 6 out of 10.
We found the Ecovacs, Roomba 690, and the Shark ION Robot all to be very reliable when connecting to WiFi. The Roomba 690 allows you to adjust the suction and view a map of the house. We found the Ecovacs and the Shark apps to be quite lacking, though you can manually control the Ecovacs through the app, as well as its remote.
The Roomba 690 will also work with Alexa, Google Home, and IFTTT, while the Ecovacs will only work with Alexa and the Shark ION only works with Alexa and Google Home. It is supremely easy to set up a schedule for each of these robots on their mobile apps and you can even use the handheld remote for the Ecovacs to set one, if you don't feel like pulling out your phone.
These robots all have some basic onboard functionality, allowing you to initiate a spot clean and send them back to their docking stations from the device itself. Rather than having buttons on the robot, the Ecovacs lets you do this from the handheld remote.
Next, the Samsung POWERbot R7070 came, earning a 5 out of 10. We didn't find connecting to the Samsung to be very reliable over WiFi, requiring a handful of resets where we had to delete the robot from the app and reconnect it. It can work with Alexa, Google Home, and IFTTT and offered average functionality, allowing you to adjust the suction and set maintenance reminders. This robot has both remote and onboard controls and you can set the schedule from both the app or the remote.
Hopefully, this has helped you decide which robot vacuum is the best fit for your needs and your budget, regardless if you are looking for the absolute best of the best when it comes to these automated assistants or if you are shopping on a tighter budget.
— Austin Palmer, David Wise, and Jenna Ammerman