Reduction in the number of routes by 68 or 27.4% (from 260 to 192 routes)
Reduction in distance by 21,248 km or 28.26% (from 75,188 km to 53,940 km)
Monthly savings for one wholesaler exceeding PLN 53,000
Customer: electrical goods wholesaler Oninen
Challenge: Proper selection of vehicles from a subcontractor to minimise the total distance
The customer has 36 branches (multiple base points) serviced by 150 dispatchers. Orders for the next day are collected the day before. On their basis routes for the next day are created (several). The customer does not have its own fleet, having at its disposal subcontractors with different types of vehicles of different load capacity (heterogeneous fleet with vehicle load capacity limitation): delivery vehicles up to 3.5t,
light trucks up to 8t,
heavier trucks up to 12t.
The distribution points (approximately 100 per day) are assigned the size of the goods in pallet units (fractional values are possible, as it is allowed to put the goods to several points on one pallet), which must be delivered to them, or possibly collected (loading/unloading). They also have defined opening times (time windows). Drivers set off from the base point with a working time limit of 9 hours 7:30-16:30. Distribution points are open between 8 am and 4 pm.
Description of the problem
The objective of the task was to optimally select the set of necessary vehicles to be rented from subcontractors in order to execute the orders in a given time. Additionally, the Customer expected the minimisation of the total distance (vehicles with lower load capacity have lower total weight and thus less road restrictions, which also influences the distance travelled) and thus the distribution costs. The optimisation process also took into account information about traffic jams (travel time depending on the road situation).
For the optimisation, the customer prepared a data sample from one month for the Łódź branch covering central Poland. The sample contained addresses of target locations, which were normalised and geocoded. The data was grouped into realised routes, which made it possible to perform a comparative analysis of the state before the implementation of the optimisation system with the results after its implementation. For this purpose, the actual routing of the original routes was determined and compared with the routes obtained during optimisation.
Resolution
The following average monthly savings were achieved for one branch:
Reduction in the number of routes by 68 – 27.4% (from 260 to 192 routes)
Reduction in distance by 21,248 km 28.26% (from 75,188 km to 53,940 km)
Particularly good results were obtained on so-called peak days when the number of points to visit was the highest (more than 120 points) – there was a reduction in the number of routes by 6-7 and a relatively large reduction in the total distance even by more than 40%.
There was an increase in the use of the largest vehicles (resulting in a reduction in the number of routes and kilometres) with a simultaneous shift in the centre of gravity from medium-sized vehicles to smaller vehicles (meaning a more optimal use of available loading space).
Assuming an average rate per kilometre of PLN 2.5, monthly savings for one wholesaler amount to approximately PLN 53,120.
For 36 wholesalers this is a monthly saving of PLN 1.9 million!