Table turnover for restaurants
Status: Planned
Most operators know their covers but not how long each table actually takes to turn, or where the minutes go between a party leaving and the next one sitting down.
How GroundMind approaches it
- You mark each real table as an area on a camera view.
- GroundMind records when a table becomes occupied, cleared and ready, from camera observations.
- Each turn is kept with its evidence, so you can check how a number was reached.
Where it stands today
Not yet producing real results. The building blocks (table areas, occupancy observations, event timeline) exist; seated, cleared and ready detection is still being validated against hand-labelled service.
What you need
- A camera with a clear view of the dining tables
- Each table set up with its seat count
- A validation period during real service
Limitations
- Event-triggered still images are weaker evidence than continuous footage for exact timings
- Cameras that only see staff areas cannot measure tables
How it will be measured
- Each real table is drawn as an area on a camera view and given its seat count.
- A turn is measured from the first observation of a seated party to the observation that the table is cleared and ready.
- Every measured turn keeps links to the observations behind it, so a person can check how it was reached.
What the pilot aims to validate
- Whether seated, cleared and ready can be detected reliably from event-triggered still images at your site.
- How closely GroundMind's turn times match a hand-labelled record of the same service.
- Which camera positions give enough view of the tables to be useful.
This capability is Planned. A pilot is a validation exercise at your site, not a guarantee of results.
Interested in validating this at one of your locations?