Dynamic routing is an order fulfilment strategy where orders are dynamically grouped to create an optimum delivery route. An optimum delivery route is one that will cost the least and ensures that the vehicle is optimally utilized.
Assuming delivery costs are broken down as below;
⎯ Fuel costs = Distance (Kms) covered * Fuel cost per KM
⎯ Vehicle costs = cost per trip * no of trips
To create optimum routes, it therefore requires one to ensure that the vehicle covers the shortest distance and carries the most weight per trip.
Remember, per trip, a vehicle will carry a maximum of its capacity (vehicles have a standard carrying capacity as per the manufacturing specifications).
To achieve a dynamic route;
Per territory, per delivery window, select a combination of orders that will cover the shortest distance but provides the highest utilization. (Utilization = Vehicle capacity (as per manufacturer)/total weight of scheduled orders. There is also a consideration of the capacity in terms of space occupied by a product.
In this case; You are only required to address one element of dynamic routing which is clustering shops (customer_id) around the fulfilment centers.
Pls submit an output and the algorithm you will use to cluster the shops. Plot the output on the map and share screenshots.
Check the sample
⎯sample_submission_file_centroids for the columns required and
⎯sample_submission_file_screenshots for screen shots of the clusters and the algorithm used.
⎯ dataset.csv – contains customer orders and delivery details, ⎯ territories_centroids.csv – contains the geo location of all territory
warehouses/fulfillment centers which will be the central place in your clusters (centroids).
