Abstract:
This research aims to determine suitable locations to build distribution centers for a specific company that distributes traffic safety products to highway offices across every province in Thailand using strategic methods. In this problem set, the predetermined location of customers and each corresponding demand were given. With customer locations spreading widely across Thailand, this research aims to find locations of distribution centers, which will result in the lowest cost to travel from the company's headquarter to the distribution centers and the cost of travelling to and from customers from those distribution centers by considering the transportation cost, vehicle cost and distribution centre rental cost. The research aims to provide the company with a comprehensive decision-making solution in location to invest in distribution centers to minimise the company's logistic expenses. Compared to the company's current delivery strategy, investing in correct distribution center locations shall positively impact the company's profitability through cost savings and improve its competitiveness in the marketplace.
The solution methodology was decomposed into two parts. The first part is clustering. This minimises the data size by grouping customers based on their geographical location through the k-mean clustering method via a GIS platform. The result obtained from this clustering process is then fed into the second part, CVRP optimisation. This optimisation method is used to find the optimal location for the distribution center and the optimal routes from that distribution center to all customers in each cluster.
This research grouped customers into four and five clusters to determine the best investment location. It was calculated that four clusters would give the company a higher cost-saving benefit. The provinces with the lowest CVRP objective cost for four clusters are Phitsanulok, Pathumthani (the headquarter), Nakhon Ratchasrima, and Chumphon. This method will decrease the company's transportation cost by 40%.
