A real store, and real customer checkout data are used to calculate customer densities at each location. Colour codes are with descending order: blue, red, purple, orange, pink, green, cyan, grey and nothing. Then using Manhattan City Block distance all locations above 0.3 (30%) of customer densities have been linked to form a most popular customer path. Once this path is formed a Genetic Algorithm will minimise or maximise this path's length.
In the same store, this time each individual customer path has been internally calculated (using a simple rule: customers move from each purchase point to the next nearest point, following their shopping list) Then all customer paths have been summed for each aisle patch, to calculate the customer path densities (bottom-up). The path densities are displayed as a relative density map (using the same colour code - blue is the highest density as above).
Demonstrates real customers moving in a real store, using timestamped checkout data (We know what each customer bought and when they checked out). Three rules determine the customers realtime attributes:
1 - A user defined parameter determines time delay is realised while purchasing each department (it is possible that wine can take 2 minutes, while milk takes 5 seconds).
2 - A user defined parameter determines the walking speed in store.
3 - A user defined parameter determines the delay at the checkout.
Using these simple rules the system works out each customer's location at any time, and the simulation runs on a PC (for 6000 plus customers)
