New optimisation methods revealed the best locations for distribution centres

Agent-based simulation and network optimisation tools can provide efficient solutions for today's complex supply chain and logistics problems, according to Javad Rouzafzoon's doctoral dissertation at the University of Vaasa. Rouzafzoon's dissertation focuses on three research projects that address common supply chain management and logistics problems.

The study used network optimisation tools to identify the top five ideal locations for a distribution centre in Finland, Norway, Sweden and the Nordic countries, based on demographic geographical information and trip distance.

– For example, at the Nordic level, the top five distribution centres include Stockholm in Sweden, Hämeenlinna in Finland, Oslo in Norway, Halmstad in Sweden and Steinkjer in Norway, says Javad Rouzafzoon.

Based on the number and location of people living within each five-kilometre square area in the Nordic countries, Hämeenlinna is one of the top five locations for a distribution centre that offers the shortest travel distance and best access to consumers.

Optimised collection points and fleet numbers cut costs in forest supply chain

In a case study on the supply chain in the forest industry, Rouzafzoon modelled the entire process from the pickup of tree logs to their delivery to collection points. Two scenarios were used; in the first scenario, logs were transported to two fixed collection points, while in the second scenario, they were transported to three train stations and a factory. 

The lowest transport costs were found in the second scenario, with six trucks assigned to facility one, totalling 920,859 euros. In the second scenario, minimal logistics costs of 755,345 euros were achieved by allocating five trucks to facility 2. In addition, Rouzafzoon introduced multiple key performance indicators to measure the performance of the company's operations, such as inventory levels, travel distance and the number of trucks in use.

Optimised fleet and scheduling enhance efficiency in collecting perishable materials

In another case study, Rouzafzoon optimised the number of vehicles in a fleet and their schedules for the collection of perishable material. He created three scenarios with a total number of trucks of two, four and six and evaluated the impact of truck arrival and departure times on fleet size and inventory levels using key performance indicators, such as the inventory level of the manufacturing unit and the number of trucks in service. 

The four-truck scenario was found to be the most efficient, with manufacturing unit inventories remaining below 45 tonnes, and all materials being delivered before the end of the week. Additionally, the fourth truck was used only three days a week.

Javad Rouzadzoon

Rouzafzoon's dissertation demonstrates how an agent-based approach can be employed to resolve supply chain and logistics problems. Agent-based frameworks are created from units called agents, which can be any item, such as trucks, facilities or orders. Agents can communicate with other agents, make decisions and set goals based on their actions; thus, every component in the supply chain can be modelled as an individual agent with its priorities and conditions. 

Rouzafzoon suggests that when dealing with complex problems in supply design, agent-based modelling (ABM) is the way to go. 

– ABM can merge a wide range of supply chain components because of the agent-based structure, says Rouzafzoon. Additionally, ABM modelling can incorporate geographical features, such as routing methods and spatial location information, making it an even more versatile tool.

Doctoral dissertation

Rouzafzoon, Javad (2023) Development of transportation and supply chain problems with the combination of agent-based simulation and network optimization. Acta Wasaensia 503. Doctoral dissertation. Vaasan yliopisto / University of Vaasa. 

Publication pdf 

Public defence

The public examination of M.Sc. Javad Rouzafzoon’s doctoral dissertation ”Development of transportation and supply chain problems with the combination of agent-based simulation and network optimization” will be held on Thursday 16 March 2023 at 8 AM at the University of Vaasa.

The public defence will be organised online (Zoom, password: 412431)

Associate Professor, Dr Youssef Boulaksil (United Arab Emirates University) will act as the opponent and Professor Petri Helo as the custos.


Further Information

Javad Rouzafzoon, email: Rouzafzoon (@)

Javad Rouzafzoon was born in Iran. He has a Bachelor's degree in Science in Industrial Engineering, System Analysis and Programming from the Azad University of Shiraz, Iran (2010) and a Master's degree in Economics and Business Administration in Industrial Management (2016). He works currently as a data scientist for a healthcare company in Texas, USA.

Did you like the article?