The Situation
Wincanton provided Opturion with a transport optimisation problem to be applied to Opturion’s optimisation software. The optimisation problem was close to real-life operation, with a number of dimensions:
- It was a multi-day, multi-depot operation;
- There was a mix of company-owned fleet, subcontractors, as well as the option to do fixed price delivery through a pallet carrier service;
- There was a potential to do dynamic staging of orders via one of a set of staging locations;
- There was a potential to have more than one driver shift for each company-owned vehicle;
- There was an option for drivers to make a multi-day trip with overnight stays.
The problem had been sent to several different transport optimisation providers and functioned as a benchmark for Wincanton to evaluate transport optimisation solutions.
Opturion did an initial optimisation in which several aspects were left out, most notably the multi-day element and the different delivery options through the company-owned fleet, subcontractor or pallet carrier. In this report, we provide a comprehensive solution to the entire problem.
Our Approach
The problem was solved using Opturion’s optimisation software.
Initial Optimisation Run
For the initial optimisation run, we used all the marginal costs (fuel cost and driver cost). However, that led to the optimiser preferring the use of subcontractors over company-owned vehicles – as they appeared to be cheaper – and only a very small part of Wincanton’s fleet were being used. A second run was performed using only the fuel costs. Upon its completion, two cost plans were obtained: one which was based on the fuel cost only for the company’s fleet, and one which included all the marginal costs.
Staging Process
On completion of the optimisation run, all the orders had been allocated. We found that the majority of orders could be allocated directly, while some of them required staging. During the staging process the order was first delivered to one of the staging locations and picked up later by a different vehicle and delivered to the destination location. For a large part, staging could be avoided by using subcontractors who were assumed to finish their run at the delivery location rather than having to return to base.
Conclusion
The problem was modelled based on the provided data and the optimisation run was performed. As a result of the optimisation, all the orders were assigned. The routes for the orders that could be allocated directly, as well as for the ones requiring staging were determined. Furthermore, the proposed plan was fatigue-friendly.
It was also possible to run the optimisation experiment longer, to further reduce the cost.
At the time, the small drop service only operated from four specified locations (this was relatively easy to model using our software). However, the solution could be further improved by having the delivery picked up from a location by the company-owned fleet, dropped off to one of the four hubs, and then having the small drop service delivery to the destination location. This could also lead to a further decrease in the costs.
The Results
Customer orders remained steady during the sales transition, and financial performance in the impacted business units exceeded internal forecasts.
Sales representatives were shifted toward the future-state, guided by the new organization structure, territories, product coverage, and compensation design in both North America and Europe.
The combined company was also on track toward full integration of sales enablement tools, including revised training programs, rationalized sales IT systems and updated sales dashboards and performance metrics.
Identified net synergies for the third year expected to exceed expectations at 1.5 times the initial target.
