Running out of time to clean with all the hustle and bustle of the holiday season? We bought the 10 best robot vacuums on the market and pitted them head-to-head in an extensive barrage of tests. After extensive research and over a year of on-and-off testing, we are confident that our award winners below are the best you can get. We compared and scored the performance of each product picked up debris from both carpet and hard floors and how well it cleaned the hard to reach areas of a room, like corners and edges. We also judge each robot on how well it navigated, how easy it is to use, and how well it did at collecting pet hair. Take a look at the complete review below to find out which robot ruled them all and which 'bot gives you the best bang for the buck.
Updated December 2017
With Shark products performing exceptionally well in our upright vacuum, we had extremely high hopes for the newly released Shark ION Robot 750, even having the potential to claim the top spot. Unfortunately, the ION robot feel exceptionally short, scoring near the bottom of the pack and leaving the Neato Botvac Connected unchallenged as the top robot vacuum around. Check out the full review below to see exactly how the Shark performed and where it got swept away by the competition.
Taking home the top score of the entire fleet of robot vacuums that we tested, the Botvac Connected by Neato handily won the Editors' Choice award for being the Best Overall Vacuum. This top-of-the-line vacuum does an excellent job of cleaning all types of floors — even fluffier, medium-pile carpet. This all-around exceptional vacuum navigates itself around furniture with ease and traverses common household obstacles, such as charging or blind cords, without becoming entangled. On top of a solid performance in our cleaning and navigation tests, the Connected does a good job of picking up pet hair and is exceptionally easy to use, even featuring wi-fi connectivity. This allows you to remotely control and program the robot through its companion app, as well as check its status and current progress. While this is one of the more expensive products of the group, it's the overall winner and the clear choice if you want the best of the best.
Want to keep your floors clean without sucking your bank account dry? The Neato Botvac D80 is the best value option of the entire group, matching the cleaning and navigation power of our Editors' Choice award winner, the Neato Botvac Connected, only forgoing the network connectivity and a few of the accessories to keep the cost down. This robot is exceptionally easy to use, with a built-in screen right on the device that you can use to start the robot and set a schedule. It retains the same systematic cleaning method as the Connected — a substantial improvement over the random or semi-random cleaning power of other models. This robot can handle multiple rooms with ease, even stopping to recharge and resume where it left off if the battery starts to run low. All in all, the D80 is the best bang for the buck that you can get when it comes to robot vacuums and is our top recommendation if you want clean floors without cleaning out your wallet.
We spent close to 2 years testing the cleaning performance of the top robot vacuums that you can buy today available on a variety of surfaces with different types of debris. as well as the ability of the robot to avoid getting stuck and successfully navigate household obstacles. We also looked at how easy it is to use these products, from unboxing and maintenance to scheduling automatic cleaning of your home. We conducted over 25 individual tests across 6 weighted metrics to get a score from 0-100 for each model.
The scores were determined by comparing the performance of each model to its peers in each of the tests. Below we go into further detail in each metric, as well as go over our best and worst performers in each category. A key point to keep in mind is that the performance of any robot vacuum can vary wildly, depending on the exact layout of the area to be cleaned. These products can't handle stairs or steps, and even the best navigators will become quickly bogged down in a cluttered home. It's necessary to "robot-proof" almost every home before any of these products really excel at cleaning them.
The semi-random path of the iLife misses significantly more than the systematic models did in the same amount of time.
This metric is summed up by one thing: If you turn on your robotic vacuum and leave for the day, how confident can you be that you will return to clean floors? Intuitively, you might put the most importance on the cleaning abilities of these products, and cleaning performance and indeed important (earning the 2nd highest weighting in our scores), but after spending two months testing these products we concluded that navigation abilities are an even more important differentiator. Why? the main point of getting one of these is to clean for you, and it can be infuriating to come home every day to find it stuck under a chair, beeping helplessly. The performance of these can vary wildly depending on the setup of your home, and it may be necessary to spend some time "robot-proofing" your home before it gets cleaned at all, but we did find some distinct differences in how each model navigated around a room.
The D75 becoming trapped and displaying an error message on a tasseled rug. Neato recommends folding the tassels under the rug prior to vacuuming.
Most homes have countless obstacles that can trip up a robotic vacuum. Items like chairs, tables cords, shoes, and area rugs are all things that can present challenges to these products and prevent your home from being cleaned, and the ability to navigate around these obstacles is what separates these products. We tested the ability to drive over a small obstacle by traversing a laptop power cord, an area rug with tassels, an extremely long blind cord, and a shoelace without causing an error. We then tested the ability to navigate by having them clean a large empty room, multiple empty rooms, and an average room filled with furniture. You can see how all of the robots scored in the chart below.
Our primary focus was seeing which scenarios required us to intervene, and then on which areas were missed and why. Based on the results of our tests, the top performer was the iRobot Roomba 980, earning a 7 out of 10 on this crucial metric. This model cleans in a systematic way but is also particularly adept at escaping from confined areas quickly and efficiently. Following the top performance of the Roomba 980, all Neato brand models (Botvac Connected, Neato D80Botvac D75, D5 Connected, and the D3 Connected) scored a 6 out of 10. These five vacuums all navigated our test rooms in essentially identical ways, and utilize a systematic cleaning method as well. This proved much more effective than the semi-random navigation style of the Roomba 650, Ecovacs Deebot, and the iLife A4.
However, the Neato models much more prone to becoming trapped for long periods of time in tight spaces. The Roomba models were the fastest at navigating out of the most confined spaces, quickly finding an exit and weaving their way through the various chair and table legs in our course setup. All Neato models lagged behind the Roombas, taking longer to free themselves, and exhibiting significantly more difficulty navigating through the furniture legs, bumping into the legs and potentially scuffing them, but still managing to free themselves.
Next, the iRobot Roomba 650, Deebot N79, and the iLife A4, all earning a 5 out of 10. The Roomba 650, with its random method of navigation, is not designed for cleaning multiple rooms, and, as expected, performed poorly in the large room and multiple room test. The iLife and Ecovacs have a similar navigation style to the Roomba 650, but proved a little less adept at escaping confined situations, almost feeling like this pair of robots were freeing themselves by chance, compared to the Roomba 650's deliberate movements. Below you can see how well the Roomba 650 can free itself from a confined space, much quicker than some of the other models.
When it came to traversing obstacles, we found the Roomba models were much less likely to be tripped up by a shoelace or rug, but could potentially damage themselves in the process. In our tests, we found the Roombas wreaked havoc when they sucked up a thick cord, usually only stopping when the wheels would no longer turn. The Neato models and the Samsung model quickly errored out if they detected any sort of tangle, requiring rescue.
Finishing out the bottom of the pack, the Shark ION Robot earned a 3 out of 10 for its poor performance at navigating.
The Shark still trying to figure its way out from under the table.
This robot uses the same semi-random navigation method for cleaning as the Roomba 650, iLife, or Ecovacs, meaning that it fares quite poorly at cleaning large rooms or multiple rooms. However, while the Roomba was quite adept at navigating around furniture and freeing itself, the Shark lacked these skills, hurting its score significantly.
The Neato D5 ready to attempt the flour torture test on shallow carpet.
This is probably one of the most common uses for a robot vacuum, and the first one that we looked at: Can these get the junk out of your carpet, so you don't have to? First, we looked at what types of carpet are commonly found in homes and apartments and the types of carpets that these products are designed to clean. We put each robot through its paces on low and medium pile carpet (leaving the high-pile, shag carpet for the 70's) with different types of debris to represent what would commonly be found in a typical household. For some background, the pile of a carpet is the part of the carpet that you walk on, and is usually grouped into low, medium, or high, and is basically how fluffy the carpet is. The fluffier the carpet is, the harder it is to clean. None of the models we tested are really meant for anything fluffier than a medium-pile carpet.
We tested with flour, rice, oatmeal, and mini-wheats and then evaluated the amount left behind after the carpet was cleaned. For our surface clean test, we measured out the debris and spread it over the same area for each robot. We then gave each one a shot at doing its best to clean on a single pass and then cleaned with a traditional, upright model before the next contender. Our final carpet cleaning test was a true torture test: 2 tablespoons of flour spread out and pressed into the carpet. You can see how the other 'bots stacked up in the chart below.
The Neato Botvac Connected, Neato Botvac D80, and the iRobot Roomba 980 were the top contenders. Both Botvacs had a tendency for mini-wheats to get wedged in the intake zone, lacking the power to effectively crush them. These Neato models both have the combination spiral brush, which we found to be much more effective than the rubber blade brush at cleaning carpets.
We found the rubber blade extractor (above) to be less effective than the combination spiral blade brush below.
These models include both types so you can customize the vacuum to suit your needs. However, we generally found the combination brush beat the blade brush in every test and aren't totally sure what we would use it for, but feel free to experiment in your home. The bottom line is that all of these will pick up most of the mess left out for them, and anything they miss will get picked up by a second pass. This reinforces the point that these are designed to be run daily, to pick up most of the day to day mess and lengthen the time between deep cleans with traditional methods.
Next were the bulk of the models, with the D3 Connected, D5 Connected, Botvac D75, and the Roomba 650 all earned a 5 out of 10. These robots all did acceptably well with the medium-sized particles but faltered at collecting flour and Mini-Wheats.
The D3 and the D5 did a particularly poor job at collecting flour from deeper in the carpet, with the D75 only doing slightly better, and the Roomba 650 topping the three of them.
The Shark came next, earning a 4 out of 10 for its performance. This model did alright at collecting the medium-sized debris, such as rice or oats, but was quite deficient at collecting flour or Mini-Wheats.
The Shark performed slightly below average for carpet cleaning.
The iLife A4 and the Ecovacs finished at the back of the pack, both earning a 3 out of 10.
The N79 left plenty of residual flour behind.
These models overall fared poorly in our carpet cleaning tests, leaving a decent amount of debris behind in every test, and in many cases, creating an even larger mess than we started with by flinging rice and oats about the room with their rotating brushes.
The Connected cleaning up flour from a hard surface.
Hard Surface Cleaning
Most homes have some area that is covered by a hard floor and will need to periodically clean that as well. Ideally, an automatic cleaner will be able to handle these types of floors, and we wanted to find out how just how well each of these models stacked up. The results are a little more spread out here, as different design elements come into play, such as side brush, main brush type, and ground clearance. On the whole, most of these will perform better on the hardwood than on carpet, great for reducing the number of times the floor must be swept. The performance of each model on hard surfaces is shown in the chart below.
Our test used flour, Cheerios, Mini-Wheats, and oatmeal for test messes. Our goal was to test how these products dealt with a concentrated mess of fine and medium particles (flour and oatmeal), whether or not they cleaned or crushed up the larger particles (Cheerios and Mini-Wheats). The concentrated mess test allowed a single pass over the debris spread in a 5-6" circle, while we counted out a set number of the larger particles and compared the quantity missed. There was a stark difference between some the different models, largely influenced by the type of main extractor each one had.
The Botvac Connected, Neato D80 and the D5 Connected lead the pack on flour, oatmeal, and cheerios, earning the top score of 7 out of 10. This trio of models have a single rotating side brush and did a great job at collecting flour and oatmeal, picking up roughly 95% of the debris. These model also did slightly better than average at collecting the Mini-Wheats, though they were prone to becoming lodged in the extractor, rather than making it to the collection bin.
The D75, and the D3 Connected all tied for the runner-up position with a 6 out of 10. These models did well, just not quite as effective at the frontrunners, and would neglect to pick up a larger percentage of the mess. These just edged out the Roomba 980 and the Shark ION, which both scored a 5 out of 10. The 980 model left slightly more residual mess, and also lacked the clearance to pick up some of the larger particles, merely pushing them around.
The before and after of the Shark during its flour test on hardwood flooring.
The ION Robot did a mediocre job at keeping hard floors clean, doing alright with small and medium-sized types of debris, but again failing to collect any Mini-Wheats at all.
The iRobot Roomba 650 did have a standout performance with the Mini-Wheats, being the only one that had the ground clearance and the extractor power to crush up and capture the pieces in its collection bin, but it wasn't enough to offset its poor performance in other tests. This model, along with the iLife and the Deebot N79 all earned a 4 out of 10. The 650 left a ton of debris behind, particularly in our flour test, and all of these robots have large rotating brushes in the front. These brushes tended to create almost as large of a mess as we started with by flinging medium and large particles away with their larger side brushes.
The D3 progressing through our corners and edges test.
Corners and Edges
We wanted to evaluate how close to a wall, and how tight of a corner these machines could clean. This category highlights some of the differences in the design of these machines, with differences in the overall exterior shape, feeding wedges, and additional brushes or suction influencing the results.Below you can see, from left to right, the undercarriages of a Roomba and a Neato model.
All models except the Samsung and the D3 have a rotating side brush intended to sweep dirt and grime away from the walls into the path of the extractor to be sucked up. None of these are perfect, and really getting the edges of your home is best reserved for the periodic deep clean, but, in our opinion, the less work it takes for the deep clean, the better. The scores for each robot are in the chart below, with the higher scores corresponding to how close the robots got into the tight corners and how little they left along the edge.
We tested on carpet and hardwood floors, running each test in a blocked off, 4'x4' pen. Each product was started in the middle of the pen, and allowed to run long enough that it either stopped on its own or had run along the entire edge. We found rice was picked up with a 100% success rate in previous tests and was selected for this test to isolate corner and edge cleaning ability. 30 g of rice was spread out evenly around the enclosure, up to 4" away from the wall for each model, and rankings were based on weighing the final amount picked up. The flour on hardwood floor test is the best visual summary of how each design fared with corners and edges.
The Neato Connected, Neato D80, D75, and D5 all performed the best, meriting an 8 out of 10 for this metric. All of these robots picked up about 80% of the rice on the hardwood, and 75% on the carpet. All of the Neato vacuums struggled a tiny bit with the corners due to their design, but still got in closer than the other models.
The iLife, Ecovacs, and the D3 earned a 6 out of 10. The D3 lacks a side brush but has a wide main extractor, and a D-shaped design to get in reasonably tight to the borders. The iLife and the Ecovacs are the only models that have dual rotating brushes, and while these did sweep in more debris in some cases, they also flung a non-trivial amount around the room, especially on the hard floor.
The Ecovacs does an alright job at cleaning along a wall, but leaves a decent amount behind in the corners.
The Shark ION, 980, and the 650 did about average, earning a 5 out of 10. The intake on both Roomba's is approximately 4" from its sides, and relies heavily on the side brush to fling debris into the path of the extractor. This seemed to be extremely hit or miss in our opinion, as it collected some of the debris, but also flung it all over the place. This was particularly pronounced on the hard floor and lessened on the carpet with the added resistance to the side brush.
The Ion robot relies heavily on its side brushes to pull debris away from edges and corners.
The Shark did alright on the hard surface with flour and the carpet with rice tests, only leaving a small border of debris uncollected. However, it did struggle with cleaning in close to the corners of the test pen. It also struggled across the board at collecting the flour from along the edge when cleaning carpet.
Senior canine tester Chewie seems bemused as the Neato D5 collects the excess fur he shed.
In the US, the ASPCA estimates that 70-80 million households have a dog, and 74-96 million have a cat. That adds up to a lot of fur that needs to get cleaned up on a regular basis. For many pet owners, this may be the sole reason you are considering getting some automated assistance in the cleanup. Then again, some of you may be here simply to find out which model is most likely for your cat to ride, and unfortunately, our tests did not provide conclusive results as to which product will perform best on that metric but leave that up to your best judgment. We tested the cleaning performance of each model on low and medium pile carpet, as well as on hard surface with donated hair from a friend's dog. For the carpet tests, we blocked off 50 sq. ft of space, spread out 5g of hair for each test, breaking up the tufts of hair and evenly pressing them into the surface of the carpet. None of the hair was placed within 7" of the end, ensuring that we were solely testing the ability to pick up pet hair, not edge-cleaning ability. The final scores for the pet hair test are shown below.
The Botvac Connected and the Neato D80 performed exceptionally well, both earning a 7 out of 10 for collecting the highest percentage of pet hair in our test. The D3 and the D5 followed closely behind earning a 6 out of 10, far outpacing both the Botvac D75 and the Roomba 650, both earning a 4. The Roomba 980 was about average earning a 5 out of 10, the Deebot N79 and Shark ION slightly less so with a 4 out of 10, and the iLife did poorly, earning a 2 out of 10. The Botvac Connected has a combination extractor brush, with plastic blades and bristles, picked up the most hair in the carpet tests, and we would estimate 99% of the hair picked up ended up in the collection bin, rather than caught in the extractor.
The Neato Connected after out Pet Hair test, showing that most of the hair managed to make it into the extractor bin, rather than becoming tangled elsewhere.
The D3 and the D5 have the identical brush as the Connected and the D80, but left a tiny bit more hair, dropping their score a point — evidence that these models have slightly less suction power. Both Roombas and the Botvac D75 fared poorly with this test, leaving behind plenty of hair, with most of the hair picked up tangled in the extractor and bin opening, rather than the collection bin. Based on our tests, the lack of bristles that penetrate deep into the carpet was the downfall of both these models. The Ecovacs collected a decent amount of pet hair, but most of it was trapped in the bristles of the brush, rather than making it into the collection bin. However, it didn't leave very much residual hair behind. The Shark had a better yield of hair actually making it into the collection bin, but left more hair behind on the floor than the Ecovacs. The iLife simply isn't suited for collecting pet hair, with the majority becoming tangled up in the brush and large amounts being left behind on the floor.
The aftermath of out Pet Hair test with one of the Roomba models.
All models picked up roughly double the amount of hair from the medium pile carpet when we repeated the test on low-pile carpet, but relative performance remained constant, with the Botvac Connected in the lead. Our results show that one of the most important considerations if removing pet hair is a major concern, and we would highly recommend looking at the longest bristle, combination extractor head that is available if you are primarily looking to clean up pet hair.
Our hard surfaces test was more of a torture test: 3 grams of hair spread out in a 6" X 18" rectangle, with a mix of large clumps and loose hairs. We really wanted to push the extractors to the limits, and see if they could handle this without tangling, and get all the hair into the extractor bin
All models did very well in this category, cleaning up all of the pet hair we laid out. The Roomba 980 was flawless, with the other models slightly lagging due to some hair caught up in the brush. We do recommend if pet hair removal is your primary goal, that the robot is run on a daily basis, and that you regularly check and clean the extractor for trapped hair, no matter which model you use.
The iLife included a handy remote for controlling the vacuum and setting a schedule.
Ease of Use
This category evaluates how easy it would be for someone to unbox and set up any of these 'bots for the first time, how much regular maintenance is required and the scheduling abilities of the robot. The entire point of buying one of these is to reduce the time and effort you spend on housework, not increase it. Ideally, we were looking for a product that was easy to set up for the first time, create a schedule to have it automatically clean when we weren't around and did not require a bunch of maintenance to keep going. The chart below shows the scores for the ease of use metric.
The 650 was the easiest to use of the bunch, earning an 8 out of 10. Scheduling was a little easier than on the Neato models, with the D75 and the D80 being the next easiest to use, both deserving a 7 out of 10. The Botvac Connected, Shark ION, and the iLife were next with a 6 out of 10, followed closely by the D3, D5, and the 980, all earning 5 out of 10.
The first thing we looked at was initial setup. All of the models we tested took less than 10 minutes to unwrap and start charging on their home docks. The Roomba 980 was the only one ready to start cleaning right out of the box, with the other ones needing to charge fully before use. After being fully charged, the remaining models were ready to start cleaning, with no additional setup. All of the app-enabled robots (Roomba 980, D3,D5, Shark ION, and Botvac Connected) required additional time to pair with a smartphone and download the manufacturer's app.
The next task was to create a cleaning schedule for each one, as manually pushing a button is exactly the type of labor you were looking to avoid by getting an automated house cleaner. The Roomba 650 was by far the easiest to schedule, with a dedicated schedule, day, hour, and minute buttons on the device itself. Both the Connected, D80, and the D75 have a scheduling section in their on-screen displays, easy enough to navigate through with up, down and back keys, with the Connected having an intuitive scheduling feature in the app. It was only possible to set the schedule of the D3 and D5 through the app, as these models lack screens on the device. The Roomba 980 could only be scheduled through the app, which could be problematic. We found it to be somewhat difficult to install and pair both of the smartphone versions initially, requiring multiple tries to be successful. The iLife and Ecovacs both came with a remote to set the schedule. The Ecovacs also has the ability to be controlled by a mobile app, allowing you to start the robot remotely and set the schedule with ease.
The Shark ION is quite easy to use, having a minimalist interface on the device itself — Home, Spot, and Dock buttons — with the majority of the interface pushed to the mobile app. The app allows you to set a schedule up without much hassle and is a breeze to use.
We found the day to day maintenance of these robots to be the comparable across all models, emptying the collection bin after each use, and cleaning the filters per the manufacturer's instructions. In all cases, these tasks took very little time and effort, certainly less time than we would have spent otherwise cleaning.
Some of the vacuums, ready to test!
Selecting the right robot vacuum can be difficult. Every home has a different layout and different floors, and there is no universally perfect robot for everyone. Trying to balance your budget while finding the perfect robot for your home may seem a monumental task, but we hope that this review has helped you narrow down the selection and make it a little bit easier to choose.
Very few people would pick housework as their activity of choice for their free time. As in no one. Ever. Cleaning your home is, unfortunately, one of those things that simply just has to be done, and cuts into time that could be spent doing more important things, like spending time with family and friends. A robotic vacuum is by no means a necessity for the home, and none of the models we looked at will entirely replace the traditional, upright type, or vacuum as well for that matter, but what they will do is reduce the time you spend cleaning.
Originally thought of as novelty items, the increased effectiveness and decrease in price make these a purchase worth considering. These are designed to do a surface clean of your house, and ideally be something that requires little assistance from you after the initial setup. These won't eliminate the need for periodic, manual, deep cleaning of your home, but they will increase the time between them, and most likely improve the overall cleanliness of your home. All in all, these aren't Rosie from the Jetsons, but they do a decent job of cleaning your house if you don't expect too much.