🔴 Website 👉 https://u-s-news.com/
Telegram 👉 https://t.me/usnewscom_channel
Why you can trust TechRadar
We spend hours testing every product or service we review, so you can be sure you’re buying the best. Find out more about how we test.
SwitchBot S20: two-minute review
The SwitchBot S20 is a hybrid robot vacuum with a particularly impressive mopping setup. While most robot vacuums feature flat mop pads that tend to spread spillages around, the S20 is equipped with a roller that’s fed a continuous stream of fresh water, while any dirt is scraped off as it goes. The model also arrives with a comprehensive dock that not only empties the bot’s onboard bin, but drains its dirty water tanks, refills the clean water tank and dries the mop pads, too.
However, the S20 gets into a pickle if set to vacuum after it has completed a mopping run, when its roller is still a bit damp. In this situation, any debris on the floor is spread around, and no discernible vacuuming takes place. That’s okay if your home’s layout makes it easy to keep vacuuming and mopping completely separate (with the mopping last on the list), but I wouldn’t recommend the S20 as a bot that can transition smoothly between the two tasks.
It’s also rather noisy. I’d say it’s louder than average in every cleaning mode, but the sound of it auto-emptying is especially alarming. For navigation the bot performed well, but did prove occasionally erratic – perhaps down to that ultra-sensitive object avoidance – and the app could be improved with a few tweaks.
At full price, this bot sits in the premium bracket, but hefty discounts are common and deliver much better value for money. If mopping is your priority and you spot a decent deal, the S20 could be a great buy.
A final note before I begin with my full SwitchBot S20 review. If you’re here because you heard that SwitchBot makes mini robot vacuums, be advised that the S20 isn’t one of those bots. The robot itself and the dock are full-sized, and also a little chunkier than average. Check out the SwitchBot K11+ if you’re after a tiny home-helper that can tuck into tight spaces.
SwitchBot S20 review: price & availability
- List price: $799.99 / £799.99
- Launched: September 2025
- Available: US / UK
The SwitchBot K20 launched in September 2025, with a list price of $799.99 in the US and £799.99 in the UK. At time of writing, there was a voucher on the SwitchBot US site to activate a 44% discount (taking it to $448) and one for 39% off on the UK sites (taking it down to £488). Given the availability of big discounts so close to launch, it seems likely that this bot will not be sold at full price very often.
There’s also an “Auto-Fill & Drain System” version of this model, which you plumb into your home’s water supply. It costs the same at ticket price.
At full price this is a premium robot vacuum, but those chunky discounts see it sit more middle of the pack. On paper, at the discounted price it’s a decent value proposition, but just make sure you’re happy to live with the various quirks before you take the plunge.
- Value for money score: 3.5 out of 5
SwitchBot S20 specs
|
Max suction: |
10,000Pa |
|
Robot diameter: |
14.4in / 36.5cm |
|
Robot height: |
4.5in / 11.5cm |
|
Dock dimensions (W x D x H): |
14.9 x 8.5 x 18.1in / 38 x 21.5 x 46cm |
|
Dust bin volume (dock): |
4L |
|
Water tank volume: |
2.5L clean; 2.5L dirty |
|
Base type: |
Charge, auto-empty dust, empty dirty water, refill clean water, dry mop |
|
Max threshold clearance: |
0.8in / 2cm |
|
Smart home compatibility: |
Matter (Apple Home, Alexa, Google Home) |
|
Mop: |
Roller |
SwitchBot S20 review: design
- Rather large, hybrid robot vacuum and mop
- Roller mop that’s continually cleaned as it mops
- Dock empties dust, dries roller and drains/refills the onboard tanks
SwitchBot is known for making tiny little robot vacuums, such as the dinky SwitchBot K11+, so I was quite surprised when I unpacked the K20 to find that it wasn’t small at all. In fact, far from being scaled-down, this bot is on the chunkier side for a modern robot vacuum.
By way of example, the Roborock Qrevo Master (a not especially swish or streamlined robovac) measures 13.9in / 35.3 cm in diameter and 4.1in / 10.3 cm tall, whereas the K20 is 14.4in / 36.5cm wide and 4.5in / 11.5cm tall.
The headline feature is the mop setup. The K20 has a roller mop that applies 1kg of pressure to scrub dirt from your hard floors. The roller is continually fed with clean water as it rolls across your floor, while a scraper siphons off the grime and dirty water and holds it in a small secondary water tank. While the roller can’t kick out to the side of the machine for edge cleaning, as some premium models offer, it’s still a very modern mop setup.
The suction stats are good, too, topping out at 10,000Pa. Until recently, that was about the maximum you’d find anywhere. Nowadays, there are premium bots that can manage twice that; but 10,000Pa should still be ample for most people’s needs. There’s a fixed side-sweeper brush and a single rubber roller.
Many modern robovac brands are exploring ways to help their bots scale tall room thresholds, but not so much SwitchBot. The chunky wheels should enable the S20 to make its way over steps up to 0.8in / 2cm – and perhaps a bit over that, based on my tests – but beware if your home has thresholds beyond that figure.
On the top side of the S20 you’ll find three physical buttons. One is your on/off, another sends the bot back to its dock, and the third I’m not quite sure about. In any case, you’ll generally be controlling your bot using the companion app. It’s also Matter-compatible, so you can control it using Alexa, Google Assistant and so on.
There’s also the LiDAR turret – or rather the LiDAR SLAM, as a label informs me. SLAM is an older, less-efficient method of robovac navigation (here’s more on LiDAR vs SLAM in robot navigation), but there’s no mention of it anywhere else in the literature. This is bolstered by an AI camera mounted at the front, for object identification and avoidance.
Image 1 of 4
The dock is designed to be able to take care of a number of maintenance tasks for you. It sucks the dust from the small onboard bin into a larger dust bag, drains the bot’s dirty water tank and refills its clean one, and washes the roller and dries it with warm air (122ºF / 50ºC).
Unlike other similar models I’ve tested, the bot here docks only a short way into the base station, rather than driving right inside. I’m not sure exactly how it dries because the roller is hidden under the bot, away from the dock.
SwitchBot S20 review: performance
- Excellent mopping and solid vacuuming, but a little noisy
- Sensitive object avoidance; occasionally erratic navigation
- Vacuuming with damp mop disastrous
Setup for the S20 was fairly standard. I performed a “quick mapping” run, where the bot ventured into all the rooms it could find, looked around a bit (without trying to clean them) and started to put together a map. No major issues here, although the bot did miss a room (I was confident it would find it later, and it did). I then edited the map and we were ready to start cleaning.
In terms of cleaning options, you can vacuum and mop together, just vacuum, or vacuum then mop. There are four suction modes (Quiet, Standard, Strong and Max) and Moist or Wet mopping modes.
Vacuum performance
To kick things off, I sent the bot off on a whole-floor vacuum, using the Strong suction mode, to get a general feel of its performance. My initial reaction was that it seemed noisier than other bots I’d used, with a vacuuming pitch that was a little irritating.
On the plus side, though, I could hear it ramping suction up and down as it moved from hard floor to carpet. When I paused and checked the onboard bin, I found plenty of dust and dog hair in there. The following auto-empty process was the loudest I’ve experienced in all my time reviewing robot vacuums.
Now on to TechRadar‘s standard suction tests. I sprinkled a mix of dry oats, lentils (to represent chunky debris) and dry tea leaves (to represent fine debris) on a carpeted floor, and asked the bot to clean the area twice, in Strong suction mode.
It performed fairly well, sucking up the lentil/oat mix efficiently, and clearing a good amount of the tea as well. I followed up with an area clean in Max mode, which removed even more tea, leaving only a light sprinkling. This is a solid result – it’s a purposefully tricky test.
Image 1 of 3
On hard floor, the test didn’t go quite as smoothly. I ran my first hard floor test after I’d used the S20 for some mopping, and it was a disaster – by the time the S20 made its proud “Cleaning complete” announcement, there was tea, oats and lentils all over my kitchen floor.
Repeating the test on a different day, with the mop roller completely dry, yielded much better results. Although the bot flicked some particles about and left some in the grouting between tiles, it cleared most of the mess on the first pass.
With most robovacs, the mop pad(s) can either be removed entirely or will lift to traverse hard floors, but there’s no such option here. It isn’t the end of the world – you just need to make sure you save the mopping for the last task of the day.
Mop performance
I was dismayed and confused to see there isn’t a “mop only” mode here, especially since it looks like this bot can do both things separately (“Vacuum then mop” is an option). The mopping here is more involved than usual – the bot isn’t just feeding water onto the mop pads, it’s also scraping them clean as it goes. Perhaps for that reason, the S20 emits a rather grating, droning noise when it’s mopping.
It also takes quite a while for the roller to completely soak through – in my test, I found it had almost completed mopping my whole (quite large) kitchen before the striped track marks became solid. It isn’t unusual for hybrid robot vacuums to take a while to get their mops properly wet – I usually like to remove the pad and wet it down myself before setting any bot out on a mopping run. That’s more of a hassle here, though, due to the way it’s attached to the bot.
While mopping tests didn’t go completely smoothly, I got good results in the end. First, I drizzled soy sauce and soy milk on a tiled floor, and asked the bot to do an area clean in Wet mode.
The first area clean was scuppered by the fact the bot seemed to identify the milk as an object to avoid. It skirted carefully around the puddle, smearing the soy sauce in the process, but not really achieving any actual cleaning.
After spreading the puddle of milk around a bit to blur the edges, it was more successful on the my second run (shown in the video below). It delivered on its promise of clearing the spillage rather than spreading it about, as can happen with flat mop pads.
Buoyed by my success, I also ran a test to see if it could clear smeared ketchup from the floor. Unfortunately, it didn’t manage to make much of a difference to the mess after one pass in Wet mode. That’s not unusual – it’s a difficult test – but I had gone in with quite high hopes for the roller mop.
xxxx does the dock dry the roller successfully?
Navigation and object avoidance
Moving on to navigation. Again, my experience was a little hit-and-miss. The SwitchBot S20 generally followed a logical up-and-down pattern, but there were plenty of occasions where it decided to – in the words of Fleetwood Mac – go its own way. It also had a little trouble clearing taller room thresholds, but managed them all in the end.
To test object avoidance, I placed a tissue box, shoe and charging cable on the carpeted floor in my lounge, and asked the bot to conduct an area clean. As you might have suspected from the its behavior in the mopping test, the S20 certainly has some object avoidance skills.
It successfully avoided both the cable (unheard of) and the shoe, and after deciding it wouldn’t fit between them, took a long route around the coffee table to complete cleaning on the other size of the requested zone. I’ve never seen that before. It wasn’t quite perfect – it did push the tissue box right across the room – but it is still very solid performance.
- Performance score: 3.5 out of 5
SwitchBot S20 review: app
- Fairly easy to navigate
- Some frustrating quirks
- Part of a general SwitchBot app, for all the brand’s products
The app is generally fine, with all the key adjustment options presented in a logical way. However, there are a few annoying quirks that stop it from being perfect.
I had a seriously frustrating time trying to divide the hall from the toilet, with the app refusing to give me the solid line required for an acceptable room division; it repeatedly bumped my adjuster off to elsewhere on the map. I managed to create a divider eventually, but it’s on the slant, and not at right angles to any wall (to be clear, this is not the situation in real life).
Maps and rooms have an eight-character limit so you can’t, for example, create a “Top floor”. Oh, and there’s no “Hall” or “Corridor” option within the room defaults – you can make one (just about, in the case of “corridor”, and that pesky character limit) but I’m baffled as to why it isn’t there in the first place.
Finally, because it’s part of a general SwitchBot app for all the brand’s products, you’ll have to click through a couple of screens before you reach the robovac options. It’s fine if you’ve bought into the SwitchBot ecosystem and have multiple products to control, but a bit annoying if you’re using the robot vac on its own.
Should you buy the SwitchBot S20?
|
Attribute |
Notes |
Rating |
|---|---|---|
|
Value |
Premium at list price but regular large discounts deliver decent value. |
3.5 / 5 |
|
Design |
Larger than average, with an advanced roller mop setup and comprehensive clean station. |
4 / 5 |
|
Performance |
Above-average mopping and decent vacuuming, but don’t try to vacuum with a damp roller. Excellent object detection. Rather noisy overall. |
3.5 / 5 |
|
App |
Includes all the key information, presented logically, but there are a few annoying quirks. |
3.5 / 5 |
Buy it if…
Don’t buy it if…
How I tested the SwitchBot S20
I used the SwitchBot S20 as my regular robot vacuum for a fortnight, to get a feel for how it performs day-to-day. I also ran a series of standardized vacuuming and mopping tests, to see how it would cope with spillages of all kinds. I dug into the app and assessed the different options there, as well as gauging how easy it would be to use and if things are presented in an intuitive way. I compared my findings to other robot vacuums I’ve tested, and weighed everything up against the cost to ascertain if the bot delivered good value for money.
Read more about how we test robot vacuum cleaners
- First reviewed October 2025