How long does a window robot take? Count the whole job
A robot’s travel time on one pane is not the time required to clean every window. A complete session includes setup, attachment, transfers, pad care, frames, touch-ups and equipment cleaning. Measure two outcomes: elapsed session time and the minutes of active work you personally perform.
In this guide
- Two clocks answer different questions
- Count panes and sides, not just windows
- A worked example with visible assumptions
- Compare models by the stage they change
- Sensitivity: when transfers become the deciding factor
- Account for overlapping work
- When is the result useful enough?
- Use the result to choose
- Next step
Two clocks answer different questions
Elapsed time runs from the start of preparation until clean equipment is put away. Active time covers hands-on tasks. If the instructions allow you to do another nearby task safely during a pass, that interval need not count entirely as manual cleaning. This does not authorise leaving the machine without the supervision its instructions require.
| Stage | Measure | Common comparison error |
|---|---|---|
| Setup | Pad, solution, power and restraint | Ignoring the beginning of the session |
| Travel on glass | The chosen mode’s duration | Applying it to different dirt |
| Changing panes | Removal, transfer and attachment | Counting area but not separate surfaces |
| Finishing | Frames, corners and touch-ups | Comparing different work scopes |
| Equipment care | Pad, tank and storage | Stopping the clock too soon |
Count panes and sides, not just windows
A window can contain several separate panes, each with two sides that might need cleaning. Six windows with two panes each, cleaned on both sides, create 24 working surfaces. Some may be inaccessible or unsuitable for the robot. Exclude any surface the instructions do not permit or where equipment cannot be handled safely.
Two homes with the same glass area can therefore need very different numbers of transfers. A large uninterrupted pane and many small opening sections are not the same cleaning job.
A worked example with visible assumptions
Illustrative model, not a device measurement: assume eight surfaces, four minutes of robot travel per surface, one minute of handling per surface, five minutes of setup and seven minutes of finishing and equipment care. A sequential session takes 8 × 4 + 8 × 1 + 5 + 7 = 52 minutes. Active handling under those assumptions is 20 minutes, before any additional active supervision.
If manually cleaning the same surfaces took 35 minutes including equivalent finishing, the robot would not finish sooner in this example. It could still reduce hands-on work. Replace every input with a measurement from your own home before using the result to justify a purchase.
Compare models by the stage they change
Different modes and stations affect different terms in the calculation. Automatic pad washing can reduce manual handling, a faster pass can reduce time on glass, and a battery station can simplify positioning. None automatically removes transfers between separate panes.
- Agree the scope: which sides, whether frames are included and what finish is acceptable.
- Time one complete session by stage.
- Record pad changes, repeat runs and touch-ups.
- Inspect after drying under comparable lighting.
- Repeat during another ordinary session before treating the result as representative.
For a fair manual comparison, include cleaning materials, access preparation and putting them away too. Compare the same standard of finish, not a quick manual wipe with a robot cycle followed by detailed hand polishing.
Sensitivity: when transfers become the deciding factor
The calculation below holds robot travel at four minutes per surface, handling at one minute per surface, and preparation plus finishing at 12 minutes in total. It is not a forecast for any model. It isolates the effect of the number of separate surfaces while keeping other assumptions constant.
| Working surfaces | Robot travel | Active handling | Sequential elapsed time |
|---|---|---|---|
| 4 | 16 min | 16 min | 32 min |
| 8 | 32 min | 20 min | 52 min |
| 16 | 64 min | 28 min | 92 min |
If handling rises from one minute to two minutes per surface, 16 surfaces add 16 active minutes. That can matter more than a small difference in travel speed. Measure removal, carrying and preparation instead of assuming each takes only a moment.
Account for overlapping work
Do not automatically add every robot minute to every human task if some genuinely happen at the same time. Measure total elapsed time with one clock. Separately record your active work intervals without counting the same minute twice. If required supervision prevents another activity, include it as committed time even when your hands are idle.
Treat frames and sills consistently in both alternatives. Comparing manual cleaning of everything with a robot cleaning only glass exaggerates its benefit. Comparing recently soiled windows with windows not cleaned for a long time also creates a different task.
When is the result useful enough?
Keep a record of the pane count, sides, soiling, mode, pad preparation and criteria for corrections. Repeating this during the next normal session shows whether the first result was unusual. You do not need a laboratory test to choose a more convenient routine; you need comparable scope and honest time accounting.
Use the result to choose
Buy for the most burdensome stage in your routine. That could be cleaning large panes or repeatedly caring for a dirty pad. “More elapsed minutes, fewer active minutes” can be a useful result if it fits your schedule and the required supervision. The shortest number on a product page is not enough to decide.
Whole-session window-cleaning calculator
Replace these example inputs with your own. Sequential session: stages do not overlap. Handling per surface includes moving the robot and pad care. Supervision is added only during robot travel; other stages already count as active work.
Session = surfaces × (travel + handling) + setup + finish. Active time = surfaces × handling + setup + finish + surfaces × travel × supervision share.
Calculations stay in this browser. This tool does not save or send the entered values.
Next step
Use the session calculator and compare stations by the work they actually reduce.
Sources and method
Documentation checked on 11 September 2026. Model information comes from the manufacturer materials below. Checklists and worked examples are RoboMorrow editorial work, not results of our hands-on device tests. Confirm functions and equipment for the exact variant and market.