Less than people fear, more than the ads suggest — call it five minutes a week. Weekly: hair off the roller and side brush. Every couple of weeks: tap out the filter, wipe the sensors. Monthly: dock bag and mop pads.

The actual schedule

Weekly, pull hair off the main roller and the side brush, and check the wheels and front caster for wrapping. With pets or long hair, do this more like twice a week.

Every couple of weeks, tap out the filter over a trash can, and wipe the cliff sensors on the underside and the LiDAR window on top with a dry cloth.

Monthly, change the dock bag if you have one, and wash the mop pads properly — most are machine washable, and none of them should go back on damp.

Every few months, replace the side brush and the filter outright. Both are cheap wear items and both quietly degrade performance long before they look worn out.

The one thing worth knowing

The number one reason people email me saying their robot “stopped picking up” is a clogged filter. Not a dying motor, not a defective unit, not a robot that was never any good. A filter nobody has touched in six months.

Airflow is what carries debris into the bin, and a loaded filter chokes it. You can have the highest suction rating on the market and pick up nothing. If your robot’s performance has dropped off, check the filter, the roller and the intake path before you conclude anything is wrong with it.

This is also why I track maintenance complexity as part of every review. A robot that’s a pain to maintain is a robot that doesn’t get maintained, and that’s a robot that stops working well in year two.

Dad’s Take

Five minutes a week keeps a robot working like new. Skipping it is why people think robots “wear out” in eighteen months.

A well-maintained quality robot should give you four to six years. Budget ones often don’t make it past two. The battery and the dock usually fail before the robot itself does.

What actually fails, and in what order

The battery goes first in most cases. Runtime drops off gradually and then falls off a cliff, and on many models the battery is replaceable if you can get the part.

The dock is often second — the suction motor in a self-empty base works hard in short bursts, and the mop washing hardware in an all-in-one dock has pumps and valves that are more failure-prone than anything inside the robot.

The roller bearings are third, and they’re accelerated dramatically by hair that never got cleaned off the axle ends. That one is genuinely in your control.

I keep over a hundred robots in long-term testing in the studio basement, running two to three times a day and deliberately skipping some maintenance to simulate years of wear. That’s where these patterns come from.

Run the parts check before you buy

This is the step almost nobody does and it matters more than most specs. Before you commit to a model, search for its replacement main roller, side brush and filter and confirm you can actually buy them today.

Some brands keep the full parts catalog stocked for years. Others quietly stop, and a great machine you can’t get a roller for is a great machine you throw away. This has caught out plenty of people who bought a well-reviewed robot from a brand that exited the category eighteen months later.

It’s also worth checking whether generic equivalents exist. On popular models they usually do, and they’re often perfectly good — I test both and list them on review pages.

Dad’s Take

Check that you can buy a replacement roller before you buy the robot. Takes two minutes and it’s saved me from recommending some real duds.

Any connected robot knows your floor plan, your Wi-Fi network and your cleaning schedule, and most send some of that to the manufacturer’s servers. Camera models can in principle capture images inside your home, and there have been real incidents.

What’s actually being collected

The floor map is the big one, and it’s more revealing than people assume — it’s a measured layout of your home including room sizes and furniture placement. Alongside that: your Wi-Fi network details, when you’re home based on when you schedule cleans, error and usage logs, and on camera models, images captured for obstacle recognition.

Most brands process obstacle images on the device and don’t upload them by default. Some ask you to opt in to sending photos to improve their AI models, and a few enable that by default and expect you to find the setting and turn it off. That’s the one worth checking on day one.

There have been genuine incidents in this category, including images from inside homes ending up outside the company that collected them, and security researchers demonstrating remote access flaws in shipped products. This isn’t hypothetical worry.

Four things to do about it

Turn off any image upload or “help improve our product” setting in the app. Look for it under privacy or data sharing.

Use a strong unique password on the account and turn on two-factor authentication if the brand offers it. A robot account is an account like any other.

Keep the firmware updated. Most of the vulnerabilities that have been disclosed in this category were patched — for people who installed the patch.

If it genuinely worries you, buy a LiDAR-only robot with no camera. You give up pet waste detection and some obstacle avoidance, but there’s far less to leak. That’s a real trade-off and a legitimate choice.

Dad’s Take

You don’t have to be paranoid, but you should spend two minutes in the privacy settings on day one. Most people never open that screen.

No, and no. On July 28, 2026 the FCC added foreign-produced “advanced robotic devices” to its Covered List. It blocks new foreign-made models from getting FCC authorization to be sold here. The robot you already own keeps working.

What the rule actually does

The Covered List is the FCC’s register of communications equipment that national security authorities have determined poses an unacceptable risk. Devices on it can’t receive the equipment authorization required to be imported or sold in the United States.

The definition the FCC used is broad. It covers mobile devices over about four and a half pounds, including the dock, that can move, navigate and avoid obstacles using firmware, AI or sensors over Bluetooth, Wi-Fi or cellular. That sweeps in most modern Wi-Fi robot vacuums, along with robot lawn mowers and pool cleaners.

What it does not do is reach backward. Robots already authorized can still be sold. Robots you already own keep working exactly as they did.

What it means if you’re buying

In the near term, very little. Current models are still on shelves and still supported.

A separate order issued the same day allows already-authorized devices to keep receiving software and firmware updates that mitigate harm to consumers — including security patches — until January 1, 2029. After that date the picture is less clear, and there’s a reasonable chance the deadline gets extended, as has happened with other categories.

The likely medium-term effect is fewer new models and slower US releases while manufacturers work out how to qualify. A couple of brands have publicly said they’ll look at manufacturing in the United States. That’s a story still developing, and I’m covering it as it moves.

Dad’s Take

Don’t panic-buy and don’t panic-sell. Your robot is fine. The thing to watch is what shows up on shelves in 2027.

It’s gotten genuinely good, with realistic expectations. A quality robot mop handles daily footprints, paw prints and light kitchen mess very well. It won’t scrub dried-on syrup and it won’t replace a hands-and-knees clean.

What robot mopping actually does well

Daily maintenance. That’s the honest category. Dust film on hard floors, paw prints by the back door, the light haze that builds up in a kitchen — a good robot mop stays on top of all of it, and the difference underfoot is real. My floors feel clean now in a way they didn’t when mopping was a weekend job I kept putting off.

What it won’t do is scrub. Dried syrup, a stuck-on spill, grout that’s gone gray — those need pressure and a person. A robot passes over the floor with a damp pad. It doesn’t lean into anything.

The dock matters more than the mop

Here’s the part almost nobody tells you when you’re shopping. What separates a mop you actually use from a mop you abandon in month two is not the robot. It’s the dock.

If the dock washes the pad with hot water and then dries it, you get a clean pad on every pass and there’s no mildew smell. If it doesn’t, you’re wringing out a dirty pad by hand, and the second that becomes a chore you stop using the feature entirely. A damp pad sitting in a dock also starts to smell within a few days, and that’s the complaint I hear most from people who bought a cheaper combo.

So if you’re going to mop, spend on the dock, not on the pad pressure spec.

Dad’s Take

If you’re going to mop, buy a robot with a hot wash and dry dock. Otherwise you’ll use the mop for three weeks and never again.

The good ones with AI obstacle avoidance, usually yes — and several brands will replace the robot if it fails. The cheap ones with no camera, absolutely not. If you have pets, do not buy a robot without AI obstacle detection.

What separates the ones that dodge it

Pet waste avoidance requires the robot to identify what an object is, not just detect that something is there. That means a forward-facing camera with an AI model trained to recognize it. A bump sensor can’t do it, and neither can LiDAR on its own — LiDAR sees an obstacle and steers around it only if the object is tall enough to register, which pet waste often isn’t.

Several brands are confident enough in their systems to offer a replacement guarantee if the robot fails at it. That guarantee is a reasonable signal, but read what it actually covers before you rely on it.

Who really needs this feature

A puppy that isn’t fully house-trained. An older dog having accidents. A cat with strong opinions about the litter box. A doggie door that lets the dog in and out unsupervised while you’re at work, which is my situation.

If any of those describe your house, treat AI obstacle avoidance as a requirement rather than an upgrade. The failure mode here isn’t a dirty floor — it’s the robot tracking it through every room it visits afterward, and then into the dock. I’ve watched it happen. There’s no cleaning that up quickly and there’s a real chance you’re throwing out parts.

This is the one feature I will not tell anybody to economize on.

Dad’s Take

With pets, this isn’t a nice-to-have. It’s the difference between a bad afternoon and a very bad afternoon.

Yes, on essentially every combo I’ve tested. Leave the mop pad or roller off the robot entirely, and switch the app to vacuum-only mode. It runs as a straight vacuum and never asks for water.

The two steps

Step one is physical: don’t install the pad or the roller. Most combos ship with the mop as a separate piece that clips or magnets onto the underside. If it isn’t attached, there’s nothing to get wet.

Step two is in the app. Every major brand has a vacuum-only or sweep-only cleaning mode. Set it there and save it as the default. With those two things done, the robot won’t draw water, won’t prompt you to fill a tank, and won’t run a mop cycle at the dock. Some models will show a reminder that no pad is detected the first couple of runs; that clears once it learns the configuration.

Why this is often the smarter buy

I tell a lot of people this and it surprises them. If you don’t want mopping, you’re usually still better off buying the better combo robot and never installing the mop than buying a true vacuum-only model.

The reason is where the engineering goes. Combo units get the newer brush designs, the better obstacle avoidance, the better navigation, and the better docks, because that’s where the manufacturers compete. Vacuum-only models are typically older platforms. Skipping a feature on a good robot beats buying a lesser robot that never had the feature.

The one thing to check before you commit: make sure the model lets you run the dock without water. A few of the all-in-one docks want a full clean water tank before they’ll let a cycle start. Most don’t. It’s worth a quick look at the manual.

Dad’s Take

You’re allowed to buy a feature you never use. Leaving the mop in the box is a perfectly good strategy.

On its normal setting, most run about as loud as a bathroom fan — you can talk over it, you can’t sleep through it. On max suction it’s closer to a hair dryer.

The dock is the loud part

Nobody warns you about this. When a self-emptying dock fires, it’s genuinely startling the first time — significantly louder than the robot itself, because it’s pulling debris up a tube at full power. The good news is it only lasts ten or fifteen seconds and it happens once or twice per run.

If your dock is going to live in a bedroom, a nursery or next to a home office, factor that in. It’s the single most common reason people move the dock after a week.

How to make noise a non-issue

Schedule it. This is the answer for almost everybody. Set the robot to run while you’re at work or out of the house and the entire question disappears. Every robot with an app supports scheduling, and it’s the setting I’d have you configure on day one anyway.

Use the lower suction modes on hard floors. Max suction buys you very little on a flat hard floor and costs you noise and battery. Let the robot boost automatically when it hits carpet instead of running everything at full power.

If you genuinely need quiet, mopping runs are much quieter than vacuuming because the suction motor is doing less work.

Dad’s Take

Schedule it for when you’re out and the noise question answers itself. That’s what I do.

Decide in this order: your floors first, your pets second, your budget third. Everything else — suction numbers, app features, dock size — is a tiebreaker, not a starting point.

Floors first, and it isn’t close

The single rule that saves people the most money is to match the machine to your floors before you look at anything else. Mostly hard floors opens up almost the whole market and makes mopping worth considering. Mostly carpet narrows it fast and makes brush design and suction matter more. A real mix means you need a robot that lifts its mop over carpet reliably, which is a specific feature and not a given.

If you start with suction numbers or dock features instead, you can spend a lot of money and still buy the wrong robot for your house.

Then pets, then budget

Pets change two things: you need a genuine anti-tangle roller, and you need a self-emptying dock or you’ll be dumping a fur-packed bin every single day. If anyone in the house has long hair, treat that the same way.

Budget comes last on purpose, because until you know what your house requires, a number is just a number. Once you know you need, say, mop lift and anti-tangle, then you can find the least expensive robot that does both well rather than the most expensive one you can afford.

Or let me do it for you

I spent a year building my quiz, and I built it out of thousands of real questions from viewers rather than out of spec sheets. It asks about your house — your floors, your pets, your dock space — and never asks you to know anything about robots. It takes about ninety seconds.

Dad’s Take

Anybody who leads with the suction number is selling you a spec sheet, not a robot for your house.

Two things matter, and neither is the suction number: a genuine anti-tangle roller, and a self-emptying dock with a sealed bag. My current pick for pet households is the MOVA V50.

The brush roll decides everything

You want a rubber or bristle-free roller with a comb built into it, designed so hair migrates off the ends instead of winding into a rope around the middle. That’s what “anti-tangle” should mean. Some brands use the phrase for a bristle brush with slightly wider spacing, which is not the same thing at all.

Get this wrong and you’ll be cutting hair off the roller with scissors every week, and eventually the bearing at the end of the roller gives out from the strain. Get it right and it’s a small wrap off the ends about once a month.

The dock is the second half

With pets, the bin inside the robot fills fast — sometimes mid-run. A self-emptying dock with a sealed bag means you’re touching it every month or two instead of daily, and the sealed part matters if anybody in the house has allergies. A bagless dock saves you the bag cost and hands you a cloud of dander every time you empty it.

I test this with a shedding dog, a doggie door, and a rural driveway that drags dirt in year round. That’s the environment these scores come from, not fibers sprinkled on a lab mat.

Dad’s Take

With pets, the brush and the dock decide it. Suction is the third thing on the list, not the first.