Example projects MSc internship

Here you can find examples of MSc internships of the Operations Research and Logistics group from the recent past. The aim of the examples is to give inspiration about possible internship projects at ORL, and about future work with a specialization in ORL. Unfold the items below to read more about the presented titles.

If you want to discuss your options for an internship at ORL, you can contact us via: education.orl@wur.nl. For more information about prerequisites and procedures, see the ORL thesis and internship page.


Smallholder farmers delivering the retailer in South Africa (WUR-FBR)

Limpopo province is one of the poorest regions of South Africa, especially the rural areas and its economy is based on exporting primary products, agriculture and tourism. Smallholder farmers are located throughout the area and they have low levels of productivity and average holdings of 1.5 ha per farmer. Production is adequate for subsistence purposes with little marketable surplus. Survival is extremely difficult in the modern agriculture economy for smallholder farmers. This research aimed to shed more light on the food chain structure in Limpopo province and to provide recommendations in order to improve their performance. Two research methods have been used: literature review and interviews from appropriate stakeholders in South Africa.

There is a problem of coordination with many small farmers, which is exacerbated by their geographic dispersion, low educational levels, and poor access to capital and information. For smallholder farmers it is on the short-term difficult to apply to the food safety standards. Together this has led to exclusion of smallholder farmers from modern value chains. However, studies show that smallholder farmers who were given institutional support and sufficient time, as well as technological (cooling facilities) and organizational arrangements (contractual arrangements with the exporters) managed to successfully participate in high value markets.

Expenses of Containers & Packaging and Freight Out at IFF (International Flavors and Fragrances Inc)

International Flavors and Fragrances Inc. (IFF) is a leading global creator of flavors and fragrances. The company needs more insight in the costing registration and pricing calculation of two cost elements: Containers & Packaging and Freight Out. This research aimed to describe the current situation and to design Excel spreadsheets that can give periodically insight in both cost elements. Different research methods have been used: Interviews, questionnaires and data analyses.

Requirements were defined to design spreadsheets for Containers & Packaging: (1) cost split per technology, (2) cost price/kg per technology and (3) insight in the global packaging portfolio. For Freight Out the design was required to (1) split costs per mode of transport, (2) calculate cost price/kg of freight charging deliveries and (3) insight in the costs absorbed in the Customer Selling Prices.

The interviews revealed that the commercial departments within IFF do not have insight in the share of Freight out and Containers & Packaging costs calculated within the customer selling prices. After implementing the designed spreadsheets for a specific year, several differences between countries in the current costing registration were found. In the future, IFF has the intention to register costs globally more conform and to (re-)produce the Excel spreadsheets periodically.

Batchsizes at FrieslandCampina Riedel: Application of EOQ

FrieslandCampina Riedel (hereafter ‘Riedel’) is a manufacturer of A-brand juices like Coolbest, Appelsientje, Dubbelfrisss en Taksi. This research, initiated by Riedel’s Logistic department, investigates the batch sizes (also called lot-sizes) for each product at the production department.  The study focussed on the main principle of optimality for lot-sizing, i.e. to find the best compromise between total setup costs on the one hand and total inventory holding costs on the other hand. There are two important issues in batch size decisions: 1) Should the batch sizes increase due to low(er) setup costs? or 2) Should the batch sizes decrease due to high(er) inventory costs? By applying the Economic Order Quantity (EOQ) formula, optimal batch sizes are calculated. In order to deal with the assumption of a constant demand, several scenarios with varying demand are tested. In general, this research shows that Riedel’s current batch sizes are close to the optima. Further research for Riedel should focus on expanding the EOQ model to the whole product assortment and comparing the results of other Operation Research models with the results of this EOQ research. These new approaches should be able to cope with additional constraints regarding restricted availability of warehouse- and production capacity.

The potential for improvement in in-company supply chains; Research into the theory and implementation of a supply chain analyses method in accordance with the Lean philosophy. (Ngenious)

This research describes the possibility of analysing the potential for improvement in in-company supply chains by providing an in-company supply chain analysis method. The method is composed for the consultancy firm Ngenious, commissioner of this research. The supply chain analysis provided, is based on literature review and exists of the following steps: determining the scope, a customer value study, a Current state Value Stream Map, a data-analysis, a Future state Value Stream Map and implementation advice. As a part of this research, the first three steps of the in-company supply chain analysis were tested in practice by means of a pilot study. For determining the scope, the customer value study and the Current state Value Stream Map necessary instructions and formats were made to test the steps in the pilot study. All three steps performed well in practice, although alterations have been made to the instructions and formats of determining the scope and the customer value study based on the results of the pilot. 

An analysis of key success criteria to reach globalization of LambWeston, regarding two international business units.

This research, in the field of operations management, is conducted for LambWeston, one of the top three frozen potato products producers in the world. LambWeston has one international business unit in Europe and one in the Unites States. These business units are selling to customers in respectively Europe, Africa and the Middle-East, and in South America, Asia and the Pacific. Urbanisation, rising incomes, changing lifestyles and expanding tourism in emerging countries are factors which lead to the globalization of the frozen potato industry and its customers. Therefore, the frozen potato industry currently invests in worldwide capacity to meet the demands of these upcoming consumption regions. The research objective is providing recommendations to LambWeston, concerning developments for the two international business units, to reach globalization. By means of literature research, information was gathered about key success criteria of corporate strategy, collaboration and coordination, which are necessary to reach globalization. Result from an observation of a meeting of the senior management of both business units, a questionnaire for employees on operational level and interviews with the management, indicate which barriers might have an influence on the development in a global organization and which developments both business units should take to develop into a global organization.

VSM: looking at the direct line feed process at Heineken

In the current economic situation, the business as usual model doesn’t work anymore. A company must aim at continuous improvement to maintain success. Companies such as Heineken extensively use improvements programs to make business processes more cost-efficient. In my research I want to improve the supply of cans to the production lines of Heineken. Heineken uses a process called ‘direct line feed’ for the supply of cans. The main objective of my research is to give advice for optimizing the direct line feed process. Steps taken in my research to realize this objective are 1) to describe the current direct line feed process, 2) identify the inefficiencies in the process, and 3) propose solution(s) for eliminating these inefficiencies by means of the Value Stream Mapping method. Identification of the losses in the process showed the need of an advanced information system, which linked the parties in the supply chain, to improve the distribution network. This information system can reduce or eliminate most losses and can cause a cost saving.

Production planning and scheduling at FrieslandCampina Thailand

In the past year shortages increased which resulted in a drop of the customer service level. The central research question of this internship is: “How to reach an improved customer demand for mainstream products?”. The internship was executed in three phases: I) product allocation, II) batch size determination, III) scheduling production. Recipes are allocated to production lines based on several constraints. The product allocation results in more balanced allocation in terms of the number of different recipes per line and the capacity utilization per production line. For each recipe to be produced on a specific production line, appropriate batch sizes are determined by applying the Economic Order Quantity model. Optimal batches with a certain batch size and frequency are scheduled into a four week production plan taking into account production constraints. By adapting a more structural and fixed approach in production planning and scheduling demand fluctuations and production delays can be handled better, which will result in the ability to improve customer demand. In order to implement the production plan for production lines 1,2 and 3 integration with the production plan for the entire Samrong plant is required. The proposed approach will actually be implemented at the Samrong plant.

Automation MPS/MRP of JD Edward 8.11 in Mentor Leiden

Currently, at Mentor Leiden, the MPS/MRP messages are generated and processed by an Excel-based model. This model is called Entitled Model, which provides the information about the kinds of materials that need to be purchased, the quantities and when. However, the processes of running the Entitled Model are done manually and are complicated. Therefore, this internship is about how to manage the master production planning and how to trigger materials requirement planning automatically and correctly in JDE 8.11. At the first place, the logics of MRP and MPS calculation in JDE 8.11 are investigated. Secondly, the data quality is improved by cleaning, re-evaluating and correcting related parameters and operation data. Finally, the time-series of MRP is validated. This internship will create the correct and validate MPS and MRP processes.

Improving the Logistic Process by Anticipating the Expected Demand

The floricultural sector is a sector dealing with high costs, small margins and a decreasing consumption of floricultural products. FleuraMetz is an exporter of cut flowers with its market in Western Europe and North America. FleuraMetz differs from other exporters, because it has its own affiliates which order at FleuraMetz. A small number of products is responsible for a large percentage of the turnover at FleuraMetz. These products are called Top Products and it is assumed that the demand for these products is predictable to a large degree. The goal of this study is therefore to use the predictability of demand to find an alternative logistic process for Top Products which anticipates the expected demand. The focus in this study is on Rosa Avlanche+. The supply chain for this product was analyzed by doing interviews and observations. Furthermore the data for this product was analyzed. Based on the analysis and the interviews with stakeholders in the supply chain alternative logistic scenarios were formulated which anticipate the expected demand rather than waiting for orders of the affiliate. These scenarios were tested in a simulation study. Results show that the alternative logistic scenarios can provide a lower average cost per stem and a higher profit for affiliates with a large enough demand.

Towards more cost-efficient Labor Hours Management (HelloFresh)

An exploratory research into the management of labor at HelloFresh hubs, approached with a focus on the role of Hub Managers and Team Leaders. HelloFresh is the leading global supplier of fresh ingredients to consumers. Its main product is the HelloFresh food box. One of the main cost components within the logistics operation of HelloFresh are the labor hours. This encompasses all the hours spent at the last mile hubs that are not standard driving hours to deliver the product to the customer. The labor hours account for 20 to 30 % of the total hours of work performed by a hub. The goal of this research is to identify short- and long-term improvement points to make the use of labor at the hubs of HelloFresh more cost-efficient. The focus was put on how Hub Managers and Team Leaders approach the management of their hub. This was achieved by using qualitative and quantitative research techniques. The qualitative approach existed of interviews and observations at several hubs and at the Transport HQ at Nieuwegein. The quantitative approach consisted of exploratory data-analysis of the historic route data and labor hours data of 2019 using excel. Literature was reviewed to explain the findings. The results show that especially the larger hubs are less efficient compared to the medium and smaller sized hubs. A table with short-, medium- and long-term improvement projects is presented.

Assessing the forecasting performance of Farm System SIMulator using multi-objective calibration techniques

Bio-economic farm models aim to simulate farmer’s behavior and to evaluate the consequences of future changes. The capacity of the model to reproduce historical decisions is crucial for increasing confidence to model’s results. In many cases, farmers decision making involve multiple conflicting objectives. To ensure a good forecasting performance multiple objectives should be taken into account. Non-interactive multi-objective calibration methods has been proposed to calibrate on historical farm decisions. The forecasting capacity of farm models calibrated based on these multi-objective calibration techniques has not been fully evaluated. Our objective is to assess the forecasting capacity of a bio-economic farm model calibrated with non-interactive methods in ex-post exercises. The model is used to simulate two arable Dutch farm types. It is calibrated for year 1999 and used to predict the changes occurred in year 2003.The capacity of the model to predict the changes was evaluated by comparing observed and simulated cropping patterns.. We conclude that

including gross margin maximization, labour hour minimization and some area of onion as the main objectives which could best simulate the farmers` decision making style resulted in acceptable forecasts (23% of PAD values).

Use of premature disposal policies in supermarkets

Supermarkets encounter waste or shortage situations daily. One reason is that they do not contemplate the perishability of the products when defining the order quantities. With a growing concern in food waste, the use of premature disposal policies could complement base stock policies.
This study works with a dynamic inventory simulation that includes premature disposal decisions and the sales of removed products to a secondary market. The simulation was constructed in MATLAB R2108a. The cost structure considers the procurement cost, waste cost, and shortage cost. With the purpose to evaluate the impact on the quality, a quality dependent demand and quality dependent order-up-to-level is formulated. The simulation runs 100 years of inventory simulation and includes the optimization of two optimal disposal quantities and a threshold, which defines when a disposal action should occur depending on the level of inventory after fulfilling demand. The demand elasticity is one concept introduced in the study where the quantity of a product demanded decrease or increase when there is a change in the overall quality or freshness of the products display on the shelves. Results showed that premature disposal policies improve inventory performance in terms of cost and waste under scenarios with a review period and lead time higher than one. It is also concluded that the maximum lifetime of the products is the main factor affecting the costs positively. The premature disposal policy that removes an age category eliminates wastage and keeps similar costs as the control scenario. For the fourth and fifth premature disposal policies using one optimal disposal value and a threshold, achieved the most substantial cost reduction when compared with the other three disposal policies.

Quality Decay Control of Conference Pear (WFBR)

Predicting the quality decay of conference pear, is essential not only for the consumers but also for the production planning, the inventory and sales strategies that retailers can use. In this internship, a decay quality model was used to predict the quality decay of conference pear based on data from experiments in different storage time. The results show that the quality decay models can be used to predict the quality decay, but the dataset to be used should be large with sufficient variability for pears in different storage periods and conditions.

Economic balance optimization of dairy farms ; A tool with relevant variables and parameters to compare the effect of a management decision on the economic balance (VgZ)

In this paper the development of a tool in Excel is described which is meant as a support for dairy advisors of Voergroep Zuid. With this tool you can compare the current and desired economic situation which is based on the management decisions of the farmer. By changing one or more variables you should be able to see what the impact is on the economic balance. The main goal is that the tool actually is used by the dairy advisors. Therefore, discussions with these advisors and interviews with dairy experts from extern are the basis for the development of and reasoning behind the tool. The tool should need only little input which is also easily available. Moreover, there should be only little output such that the results are easily interpretable. Otherwise, the chance is high that the tool will not be applied.