Solutions · In-house product

Afluxion Retail Flow Intelligence

Afluxion's first solution is a platform for supermarkets and cash-and-carry stores that shows how customers move through the store: footage from the cameras a store already has becomes flow indicators, a 3D Digital Twin and layout scenarios.

In development · the pilot is the next milestone

Illustrative data, to demonstrate the interface

What goes in and what comes out

In goes recorded video, exported from the store DVR or NVR. Out come totals by period and department, each with its origin recorded: which camera, which model version and which calculation rule.

Indicators

Counting

Entries, exits and customers per hour, counted when a person crosses a line marked at the entrance, in each direction.

measured

Dwell

How long people stay in each department: mean, median and the range most of them fall in.

measured

Circulation

Flow, density and speed, measured in metres on the store floor plan.

measured

Journeys

Frequent routes, transitions between departments and coverage.

partial

Forecasts

Flow and crowding forecast for the coming hours or days, compared with a recorded baseline.

predicted

Anomalies

Deviations from the expected pattern for the day and hour.

inferred

Digital Twin and simulation

3D Digital Twin

The store in 3D with heat layers over the floor plan, aggregated trajectories, replay and department selection.

Layout simulation

Today's store compared with a layout change on the Digital Twin, always marked as simulation.

simulated

Who it is for

Supermarkets and cash-and-carry stores with 1,000 to 3,000 m² of sales area and 8 to 32 installed cameras. Each team uses it differently:

  • Leadership: compare periods and stores.
  • Store managers: peaks, queues and low-traffic departments.
  • Operations and trade marketing: assess exposure and physical changes.
  • Data team: detail and exports.

How we read the numbers

Every number says which period it came from, who was counted and how it was calculated. Four seals accompany everything the platform shows:

  • Measured: computed directly from observed data.
  • Inferred: the output of a model over the data, with stated confidence.
  • Predicted: a model projection for a future window.
  • Simulated: a layout scenario, with assumptions and model version.

When quality or coverage falls below the minimum, the indicator shows as partial, with the valid period. The platform never turns a hypothesis into causation: a simulation projects circulation under assumptions, it does not promise sales.

Product status

Afluxion Retail Flow Intelligence is in development. Capabilities ship in phases, each with a measured quality criterion before moving on: counting, flow indicators, journeys across cameras, models, Digital Twin, simulation and, finally, operation with live cameras. The next step is a paid pilot in a partner store.

See how the pilot works

Conversation

Tell us what you want to solve

Tell us about the operation, the process or the store. We reply within two business days with next steps, no commitment.