How to choose a robot vacuum
The number on the box — usually a suction figure in pascals — is the least useful specification in the category. What determines whether you keep using the machine after month three is navigation, brush design, and how much of it you have to clean by hand.
Why the Pa number misleads
Pascals measure the pressure difference the fan can generate at the inlet, in isolation. It tells you nothing about airflow, sealing, brush agitation, or how the machine behaves once the bin is half full and the filter is loaded. Two vacuums with the same rating can perform very differently, and there is no standardised test that manufacturers all follow when publishing the figure.
Treat it as a very rough tier indicator — a 2,000Pa machine and an 8,000Pa machine are genuinely in different classes — and ignore differences of a few hundred between models.
What to look at instead: brush type, navigation system, and side-brush geometry. Those determine whether debris gets lifted into the airflow in the first place.
Navigation: the thing that decides whether it works
Random-bounce (no mapping)
The cheapest machines simply drive until they hit something, turn, and continue. Given enough time they cover a small room reasonably well. They cannot be told to clean one area, they cannot resume after charging, and they will take a very long time in anything larger than a studio.
Gyroscope and camera-assisted
A step up: the machine drives in rows and holds a rough sense of orientation. Coverage is much better than random-bounce. Mapping is usually not persistent between runs, so zone cleaning and no-go zones are limited or absent.
LiDAR (spinning laser turret)
Builds an accurate, persistent map. This unlocks the features people actually want day to day: clean-this-room commands, no-go zones drawn on a floor plan, resuming exactly where it left off after recharging, and multi-floor maps. It works in the dark, which camera-only systems do not.
The trade-off is the raised turret, which adds height — worth checking if you need it to fit under a low sofa or kickboard.
Obstacle avoidance is a separate system
Navigation and obstacle avoidance are not the same thing. A machine can map your home perfectly and still drive through a phone charging cable or, memorably, pet waste. Structured-light or dual-camera obstacle systems are meaningfully better at this. If you have pets or children who leave things on the floor, this is worth more than any suction upgrade.
Brushes: rubber versus bristle
If anyone in the household has long hair, this is the specification that determines how often you will be crouched on the floor with scissors.
| Bristle brush | Rubber / silicone roller | |
|---|---|---|
| Carpet agitation | Good | Good |
| Hair tangling | Tangles readily | Much less; some designs are near tangle-free |
| Hard floor debris pickup | Can scatter light debris | Generally better |
| Cleaning the brush itself | Frequent, fiddly | Infrequent, easy |
Dual counter-rotating rubber rollers are the current best answer for hair. Single rubber roller is a solid middle ground. Bristle-only is increasingly a budget-tier signal.
Mopping: understand what you are getting
"Mopping" covers three quite different things:
- Drag-a-damp-cloth. A water tank drips onto a microfibre pad the machine drags behind it. It handles dust and light footprints. It will not clean a dried spill.
- Vibrating pad. Adds high-frequency scrubbing. Noticeably better on dried-on marks.
- Spinning pads. Two rotating discs with downward pressure. Closest to actual mopping. Usually paired with pad-lifting so carpets stay dry.
If the home is mostly carpet, mopping hardware is weight and cost you will not use. If it is mostly hard floor, it is the feature most likely to change your routine.
Docks: what you are actually buying
Self-empty docks move the emptying chore from every run to roughly every one to two months. That is a real quality-of-life change. Be clear on the ongoing cost though: bagged docks need proprietary bags, which are a recurring expense; bagless docks avoid that but you empty a canister and deal with the dust yourself.
Docks that also wash and dry the mop pads solve the genuinely unpleasant part of robot mopping — a damp pad left in a dock develops a smell within a day or two. They also need plumbing or manual tank refills, and they are large. Measure the space before committing.
The maintenance nobody mentions
Every robot vacuum requires ongoing work regardless of dock type:
- Side brushes bend and need replacing, typically every 3–6 months.
- Filters need rinsing or replacing on a similar cycle.
- Sensors and charging contacts need wiping, or the machine starts behaving oddly.
- Main brush bearings collect hair even on tangle-resistant designs.
Before buying, check that consumables for the model are actually available where you live and what they cost. A cheap machine with unobtainable filters is a twelve-month product.
Before you buy: a checklist
- Measure the lowest clearance you need it to fit under, and compare against the model height including any LiDAR turret.
- Decide whether you need persistent mapping — if you want zone cleaning or no-go zones, you do.
- If anyone has long hair, prioritise rubber rollers over bristle.
- If you have pets or a cluttered floor, prioritise obstacle avoidance over suction.
- Match the mopping type to your actual floor mix, not the marketing.
- Check dock footprint against the space you have, including door clearance.
- Price out a year of consumables — bags, filters, side brushes, mop pads.
- Confirm the app does not require an account region you cannot use.
A note on our method. This guide is built from manufacturer documentation, teardowns and long-term owner reports rather than in-house lab testing. See How We Pick.