Case Article—Locating a Truck Terminal in Texas

1. Introduction

The main purpose of this case is to introduce the decision-making process regarding location planning in a real-world environment. Most textbooks in operations and supply chain management cover various decision frameworks and mathematical models (Chopra et al. 2013, Murphy and Knemeyer 2018, Heizer et al. 2022) to identify the suitable/optimal location. In practice, however, there are many nuances in location planning: from identifying selection criteria and collecting relevant data to finalizing the decision framework, thus making any optimal/mathematical solution suggested from the analytical approaches far from certainty. This happens because the location decision is dependent on many dynamic environmental factors that could be qualitative in nature and keep changing over time (e.g., supply/demand conditions, workforce, cost of doing business, and competitive landscape). A vivid example of this challenge is evident in the following case involving Amazon.

In September 2017, Amazon announced its request for proposals for a second headquarters, a $5 billion project with 50,000 employees. It received bids from 238 locations across North America (Cohn 2019). To evaluate these bids, the team at Amazon analyzed a significant amount of data from various sources, such as CNBC’s America’s Top States for Business studies, to finalize 20 candidate locations for the project in January 2018. There were several criteria considered to construct this list, including infrastructure, economy, business-friendly environment, quality of life, and most importantly, workforce. During the selection process, Amazon slowly realized that its initial hiring plan of 50,000 tech employees was infeasible. As a result, in September 2018 (three months ahead of its self-imposed deadline for the selection), the company decided to split up the project. The analysis of the 20 finalist locations led the team to choose National Landing, Virginia and Long Island City, New York. Unfortunately, the unexpected strong opposition from the people in New York forced Amazon to drop its Long Island City plans altogether and spread its 25,000 positions among the company’s 17 existing tech hubs around the country. As a result, the winner for Amazon’s second headquarters was Virginia.

The Amazon example highlights the uncertainty and dynamics of location planning in practice. This reality generally gets little attention in operations/supply chain management textbooks that tend to focus on technical aspects of specific decision frameworks (e.g., finding an optimal solution for a given list of concrete/measurable selection criteria). Our case aims to address that gap through an actual business challenge at a trucking company that was seeking to locate its new and second terminal in the state of Texas. Rather than an optimization/mathematical problem, the case focuses on the challenges of operationalizing selection criteria, collecting data, and human subjective evaluation during the decision-making process. The case targets business students, especially undergraduate students in an introductory course in operations, logistics, or supply chain management.

Our case depicts a real company, CBT (the company name has been changed for confidentiality purposes), a multidivision (dry van, refrigerated, and flatbed) freight transporter with annual revenue in excess of $1 billion. With more than 5,000 tractors and more than 13,000 trailers, CBT is one of the largest privately owned trucking companies that can provide freight service to any location within the continental United States. To serve its operations (and drivers), CBT maintains a network of truck terminals located throughout the United States. In recent years, CBT has been experiencing an increased freight demand in the state of Texas, causing an overcrowded facility, and thus increasing waiting time for service needs at its current terminal. In response to the trend, CBT explored different options, from expanding its existing facility to acquiring a new terminal at a different location.

The case primarily focuses on identifying the second and new terminal in Texas. Through assigned questions and activities, students learn to determine which potential areas to consider for the second terminal and later to evaluate them under different scenarios. Particularly, this is a two-step case that can be implemented together or separately. In the base case, students use the data in the case to select a particular zip code area while in the case extension they research specific sites (smaller geographic areas) through collecting external data relevant to these sites. Their proposal is created from insights derived from data, as well as strategic considerations for the long-term success of the business.

Our article is organized as follows. In the next section, we elaborate our objectives when teaching the case, followed by a discussion of the literature. We then describe how data are obtained for the case and the additional settings for the case extension. Preliminary student feedback on the case implementation is presented. Finally, we conclude the paper with a discussion on our teaching experience and recommendations for adaptation.

2. Teaching Objectives

The idea for this case was initiated from a real business challenge in a case competition hosted by a Midwestern university in the United States (Benson and Chau 2022). Since then, the faculty of the Supply Chain Management program has collaborated with the participating company and developed the case based on their experience with stakeholders in the project. This case can be used to achieve four objectives as outlined in the following.

2.1. Objective 1: Problem Articulation

The students assess the situation of CBT and the need to expand operations in Texas through the current terminal or a new and second one in a different area within the same state. The case primarily focuses on the process of selecting the second terminal. Students learn various criteria used to identify and evaluate different candidate locations as well as the ambiguity of the selection process due to different perspectives of stakeholders involved. This objective can be achieved in the base case through the following assignment questions. (1) What was the situation in the case? (2) What were the options to solve the problem? (3) What process should be used to construct the solution?

2.2. Objective 2: Evidence Gathering and Data Interpretation

The students identify the evidence (or data) that helps them to understand important criteria used to select the terminal and to follow the arguments made by different stakeholders in the case. There are different exhibits in the case summarizing the data linked to selection criteria, such as proximity to primary roads, freight flows, and driver distribution. In the base case, students learn to interpret data and consider trends and patterns across exhibits. The purpose of this exercise is to form an initial set of potential sites (i.e., three-digit zip codes) for evaluation purposes. In the case extension, students are required to investigate selection criteria that are external to CBT and gather data from publicly available databases. Table 1 provides assessment questions recommended for the case.

Table 1. Objective 2 Assessment

2.3. Objective 3: Concept Application

Using available (or collected) data, students apply different concepts linked to location facility topic covered in the course material to develop their recommendation for the second terminal. This requires students to use (1) optimization modeling to evaluate distance-related criteria in the base case, and (2) multicriteria decision-making framework in the case extension (e.g., the factor-rating method). The instructor needs to decide which program or programming language will be used by the students prior to the case delivery. A template is provided to help solve the optimization problems using Open Solver (an Excel Add-in); however, any open-source application, such as Python, can be used to solve the linear programs. Table 2 provides assessment questions used for this objective.

Table 2. Objective 3 Assessment

2.4. Objective 4: Reaching a Conclusion

Students are expected to reach a conclusion on where to locate the second terminal. This step requires them to critically reflect on how the solution is derived and how sensitive their solution is to different conditions/arguments. Furthermore, they should be aware of the strategic implications of their decision for the business. Finally, students present their solution and respond to follow-up questions from the audience. The discussion can be organized around the following questions. (1) What criteria have been used to reach the conclusion? (2) What assumptions have been made? (3) What could be the risks? (4) How to mitigate those risks?

3. Literature Review

Facility location is a long-term decision that has strategic implications for a business. A good location can help the firm be responsive to customer demand while keeping its costs in control (Chopra et al. 2013). Nevertheless, selecting the right location is challenging because an optimal location is dependent on many dynamic environmental factors. Furthermore, the development of a facility is capital intensive and time-consuming; thus, it cannot be economically relocated or abandoned in a short period of time. As a result, determining the best location is an important strategic challenge. In what follows, we discuss the positioning of our case in the education literature, followed by a summary of the academic literature on facility location.

3.1. Teaching Facility Location Through Cases

There are a number of cases about facility location available through various publication sources. Our review of cases from two important sources, INFORMS Transactions on Education and Harvard Business Publishing, revealed two orientations: (a) an emphasis on modeling techniques and graduate education and (b) a well-defined location problem with a given set of potential locations for selection. In contrast, our case is designed for business undergraduate students to familiarize themselves with location planning process in a team setting. Furthermore, in the base case, potential locations are not predefined; students need to identify those and support their selection with relevant data. Finally, through the base case and the extension, students get exposure to the multistage nature of location screening and the need and challenge of collecting and synthesizing data from internal and external sources for the location decision. In what follows, we review various cases relating to facility location.

Many cases in the literature have been designed to teach modeling techniques (Carraway and Hosler 1991, Klingenberg 2012, Chatterjee and Dhaigude 2017, Brusco 2022). Brusco (2022) introduces students to optimization through discrete facility models. Using the context of locating emergency medical services vehicles in Austin, Texas, the author provides an intuitive approach to solving discrete optimization problems. Our case also uses a discrete model; however, distance is just one of many criteria to consider when selecting the location. Klingenberg (2012) tasks students to find the optimal location for a franchised ice cream parlor using various models such as regression, simulation, and decision tree. This highly hypothetical business scenario has been designed to assess students’ ability to select multiple models when making decisions for the given data in the case. In contrast, our case is built on a real business challenge that generally has insufficient external data. Chatterjee and Dhaigude (2017) focus on selecting the best location for a new restaurant through developing a goal programming model (i.e., to select the location with the minimum total deviations from the set goals). Multiple criteria are used to finalize the location from a given list of potential locations. Carraway and Hosler (1991) aim to teach linear programming through a distribution/location problem with the objective to either minimize total cost or maximize contribution. Both cases from Chatterjee and Dhaigude (2017) and Carraway and Hosler (1991) provide needed data to make model-driven decisions. Our case is different in the sense that we require students to operationalize various selection criteria and consider different data sources to support their decision.

In addition to teaching modeling techniques, there are other cases that further discuss strategic aspects of facility location (Köksalan and Salman 2003, Keskin and Barbee 2021). Köksalan and Salman (2003) discuss strategic planning at a beer producer and distributor in Turkey. The location decisions are part of capacity expansion considerations. Factors to finalize these locations were briefly mentioned (i.e., water quality, demand concentrations, and proximity to plants); however, they played no role in the decision-making process in the case. Keskin and Barbee (2021) focus on the integration of location decision, inventory, transportation, and warehousing to create an omni-channel supply chain. Warehouse locations were mainly based on the preexisting networks after a merger (with an option to modify leased facilities) to serve forecasted space and storage requirements. In essence, this case considers locations (or hypothetical new ones) solely for capacity planning purposes. Our case does not entail capacity planning.

The most closely related instance to our case is John et al. (2018). This case illustrates a relocation situation in which a new location and preliminary layout are considered for a facility. Multicriteria decision making is used to identify the optimal location using factor-rating method; data on potential sites and importance weights of selection criteria are provided in the case. Our case, however, is much less structured and requires students to deal with the ambiguity and multistage nature of location screening.

Overall, our case complements the education literature by exposing students to the practicalities of location planning. Although we ask students to explore different models through assignment questions, the case focuses more on making decisions through data interpretation (vis-à-vis model-driven decisions), a highly valuable issue in practice that tends to receive little attention in the operations/supply chain curriculum (Zhao 2022).

3.2. Background on Facility Location

From an academic perspective, facility location (or location science) is a “branch” of operations research and management science that focuses on positioning one or more facilities in a given space to achieve at least one objective determined by the decision maker, such as to minimize cost/travel distance/waiting time, or to maximize profit/revenue/service and coverage/market shares (Farahani et al. 2010). Location problems can be classified in several ways, using such dimensions as decision space (continuous versus discrete), location objective (single versus multiple criteria), and uncertainty (deterministic versus stochastic). For example, in continuous problems facilities can be placed anywhere on the plane or network, whereas in discrete problems, there is a limited number of eligible locations for selection (ReVelle and Eiselt 2005). Although location problems historically use single criterion to determine the optimal solution, multicriteria decision making has become increasingly popular thanks to the recognition of the complex nature of location decision in reality (Farahani et al. 2010).

Relevant to the case study, the following section focuses on the decision-making framework, and discusses detailed analysis on distance-related criteria used to determine desirable locations. A comprehensive review of modeling techniques for facility location (Owen and Daskin 1998, ReVelle and Eiselt 2005, Farahani and Hekmatfar 2009) is outside the scope of this paper.

The location selection process typically involves several layers of screening, with each layer becoming a more detailed analysis of a smaller number of areas (Murphy and Knemeyer 2018). These layers tend to be specific to the nature of the location problem. For example, Fernández and Ruiz (2009) used three layers in assessing sustainable industrial areas: geographic area selection, selection of suitable areas, and evaluation of specific zones. Each layer is composed of several criteria used for screening. The evaluation process usually involves both analytical and qualitative techniques to determine the most favorable geographic location(s). Qualitative techniques are based on observation, opinion, and judgement, while analytical techniques are related to modeling, formulating, and solving mathematical problems. For instance, the analytical technique (if deployed) in the case study results in an integer programming problem.

When there are multiple stakeholders and criteria involved in location selection, a six-phase decision-making process may be used (Liang et al. 2021). First, the scope of the study is defined and potential alternatives within that scope, along with experts from the stakeholders, are identified. Second, criteria for the selection procedure are constructed from existing literature or expert recommendations. Third, measurement of each criterion is defined, and data are gathered for alternative locations. Fourth, the importance of criteria (i.e., criteria weights) is determined by the decision maker. Fifth, a score for each alternative is calculated using the collected data on the criteria and their weights. Finally, the resulting score is used to select the most desirable location.

More specifically, we refer students to the factor-rating method covered in the class material. This method is commonly used to compare several locations along a number of qualitative and quantitative factors (Wisner et al. 2019) through five steps: (1) identify selection criteria; (2) assign criteria weights that sum to one; (3) determine relative performance scores that vary from 1 to 100 (other scoring schemes can be used); (4) multiply the factor score by the weight associated with each criterion and sum the weighted scores across all criteria; and (5) select the location with the highest total weighted score. It is important to note that when using this case, we only introduced this method briefly to students in the classroom and illustrated it with an example in Wisner et al. (2019). As a result, we do not assess if the students strictly follow this process (readers that desire to do so can easily develop an assessment instrument using our framework presented here). Students, however, are required to measure various selection criteria and justify their decision using these criteria.

In the case study, measuring the selection criteria requires access to data from both inside and outside of the business. A discussion on criteria operationalization prior to the case delivery may be beneficial to students; criteria description can be found in the literature (Battal 2020, Liang et al. 2021). Although freight volume and driver distribution are provided in the case, other selection criteria such as traffic congestion, land cost, and legal climate require students to research publicly available databases. At a minimum, students should take advantage of available data in the case, particularly using distance-related criteria to evaluate potential areas for the terminal in the base case. In fact, distance is a popular criterion used in many classic location models. Hence, a discussion of the P-median problem (introduced by Hakimi (1964)) prior to delivering the case to students or reading relevant material (Watson et al. 2013) can be useful.

The goal of the P-median problem is to find the location of P facilities to minimize the total weighted travel distance between customers and terminals. The general problem can be written as the following integer linear program (Owen and Daskin 1998):

• i = index of customer node,

• j = index of potential terminal site,

• N_i = number of customers at node i,

• Dist_ij = distance between customer node i and potential terminal site j, and

• P = number of terminals to be located.

Decision variables are as follows:

• X_j = 1 if a terminal is located at potential terminal site j, otherwise X_j = 0; and

• Y_ij = 1 if customers at node i are served by a terminal at node j, otherwise Y_ij = 0.

Objective (1) addresses the goal of the problem. Constraint (2) requires exactly P terminals to be located (or open). Constraint (3) ensures that every customer is served by a terminal. Constraint (4) is used to tie Constraints (2) and (3) together to ensure customers are served by the terminals which are actually used (or open). Constraints (5) and (6) are binary requirements for decision variables. It is important to note that the previous problem does not consider the current network in the case study (i.e., there is an existing terminal in Texas).

To solve the previous problem, we use Open Solver, an Excel Add-in (Mason 2012). Because of the large number of decision variables (ranging from 350 to 490 binary variables depending on the criterion selected) involved in the linear program, Excel solver cannot handle the problem (standard Excel solver has a limit of 200 decision variables). Open Solver is available for download at https://opensolver.org. Instructors can use any open-source programming languages such as Python if they choose to do so.

Because of the nature of our problem setting in which two optimal locations in Texas must include an existing one, it is possible to solve the problem in our case without using integer linear programming. One can refer to the exercise in Brusco (2022) to conduct a location search via enumerations of all possible locations. Nevertheless, integer linear programming is important for larger applications and extended problems that include variable and fixed costs in the objective and capacity constraints.¹

Overall, the case provides students with the opportunity to analyze a real-world facility location problem. Furthermore, it exposes them to the complexity of the location decision, which is caused by the influence of many selection criteria and multiple stakeholders involved in the process. Despite the sophisticated modeling techniques available in the academic literature, leading the selection process in practice can be tricky due to several issues such as the large number of factors to consider with many being qualitative in nature, required consensus with management and stakeholders involved, and bias judgement throughout the process (Deloitte).² With the case, students learn to (1) deal with the ambiguity, (2) make assumptions and justify them, and (3) identify and manage risks associated with their solution.³

4. Data

The accompanying data set is organized into four Excel worksheets. The first three contain data in three exhibits in the case, including the freight flow in Texas, freight volume and driver distribution by three-digit zip codes in Texas. The last worksheet contains the distance matrix for these three-digit zip codes.

The freight flow and volume are derived from CBT’s recorded transactions in a select year for the Texas market. The driver distribution is constructed from CBT’s driver database. The distance matrix contains estimated distance in miles. In general, distance can be estimated using street network (via tools such as ArcGIS or Google Maps API) or straight-line distance. In the case study, straight-line distance is estimated using the longitude and latitude of two points (a and b) through the formula provided in Simchi-Levi et al. (2007). Equation (7) does not take into account the curvature of the earth but is accurate enough for short distances. Further discussion on distance calculation based on zip code can be found in Huggins (2019).

5. Case Extension

The extension of the case allows student to investigate selection criteria that are external to CBT (i.e., the business in the case study), including the following:

• Distance to Freightliner dealerships (used when tractor warranty work is needed),

• Traffic congestion,

• Driver safety (area crime rate),

• Proximity to skilled workforce (for the facility),

• Land cost to build the terminal (approximately seven acres),

• Taxes, utilities, and other costs of ownership, and

• Legal climate toward trucking companies by the local jurisdiction.

Three specific sites are provided for this task: (1) Houston, Texas 77060; (2) Houston, Texas 77028; and (3) Conroe, Texas 77385. They are indeed the locations considered by CBT. The location information is provided after the submission of the base case. Students are required to identify different ways to measure the criteria, collect data from publicly available databases, and finalize their recommendation for the most appropriate site.

6. Classroom Experience

This case study is primarily designed for undergraduate students in the fields of logistics, transportation, operations, and supply chain management. Prior to the case assignment, students should have a basic understanding of concepts related to facility location and some exposure to the chosen software tool used for the location analysis.

We used this case with the extension in a semester course in which the majority of students had a background in business administration. All students were familiar with Excel; hence, we chose to use Open Solver, an Excel Add-in, to perform the optimization program. We provided the students with the case and accompanying data file. In addition, a walk-through demonstration of software tools was provided in class. We have implemented the case twice in Fall 2021 and Fall 2022 in an introductory course to logistics. Using the student feedback from the first time, we have slightly modified the implementation of the second time (see a discussion on case adaptation later).

The case was assigned to teams of no more than four students per team. They were given two weeks for the base case and then one week for the extension. Following the written submission, final presentations were given in one 90-minute class session in Fall 2021. In addition to the students and the instructor, invited guests included one faculty from the Supply Chain Management program and four company representatives from CBT. Each student group was given 10 minutes for their presentation followed by a question and answer period. They were required to focus on problem description, the process and analysis to derive the solution, assumptions made, and risks associated with their recommendation.

Students responded positively to the opportunity to work on a real-world problem. The involvement of invited guests in the final presentation has made the conversations much more dynamic. The students seemed to enjoy interacting with company representatives and listening to their sharing of real challenges faced as the project was progressing. After the project was completed, students were surveyed on their learning experience through four questions. In the first two, five-point scale questions regarding students’ perceived increase in their knowledge about facility location and collaboration were asked. The latter issue was included because we expected students to collaborate with others in a team setting to finalize the location. Specifically, when using the factor-rating method, students on a team needed to agree on the importance and rating of selection criteria. The last two open-ended questions asked students about (a) aspects of the project that enhanced their learning and (b) recommendations that could help improve the learning experience of future classes.

Our student survey was primarily used for improvement purposes; therefore, completing it is not a requirement of the case assignment. Because of the small size of our classes, student feedback is relatively limited. In particular, there were 36 responses (out of 46 students) across two class sections. Ninety-five percent of respondents either agreed or strongly agreed that their knowledge about facility location has been increased after completing the project with the remaining (6%) being neutral. Regarding collaboration, 89% of respondents perceived an increase in their knowledge, whereas half of the remaining were either neutral or disagreed. Although we did not teach how to collaborate with others on a team in our classes, we have set up the case as a team endeavor. In fact, the case narrative illustrated team interactions when approaching and solving the location problem. We hoped that would be an exemplar of teamwork for students to follow. Hence, we expected students to be more collaborative when solving the problem. For the third question, different aspects of student learning were reported, including the significance/importance of location decision, the process involved, the data collection on different selection criteria, and the collaboration with others to finalize the decision (Table 3 provides a sample of student comments). In the last question, students mainly recommended changes in the implementation of the case, such as the time given, classroom interactions, and technical support. We adopted some of them in the second section; a discussion is provided in the next section.

Table 3. A Sample of Student Comments Regarding Learning Aspects

Overall, our classroom experience has been positive. Apart from the feedback data, student participation was very encouraging as they perceived strong connections to the subject matter. We have seen many more questions, comments, and interactions during the class time than prior to the case implementation. That said, our data collection was preliminary and should be improved to generate more data on specific aspects of student learning. For example, students can rate their learning on different areas of location-related decisions such as selection criteria operationalization, data gathering, data analysis, data interpretation, and use of decision support software. Such level of information may be more accurate in evaluating perceived learning and could be used to compare with the team’s actual performance that is obtained through the assessment questions provided in the case.

7. Discussion and Conclusion

The purpose of the case study has been to introduce students to a real-world problem related to facility location. Students not only learned to apply the basic concepts but also experienced the complex nature of strategic decisions in a business. The case can be tailored to different teaching scenarios while accommodating diverse student backgrounds. In this section, we articulate our experience in preparing for the case and suggest different options to use it in the classroom.

7.1. Case Preparation

This project was assigned to students in the later part of a semester course that focused on the introduction to logistics and transportation. There was no specific prerequisite for the course, except junior standing. Prior to solving the case, technical training in Excel had been provided. When we first implemented the case, students had the opportunities to use PivotTable in Excel and visualize data in Tableau in the earlier part of the semester; these were unrelated exercises, but their outcome linked to most of the exhibits in the case. We hoped that working with the transactional data earlier on could increase clarity when students got exposure to the case exhibits. Through conversations with students, we received mixed responses; some appreciated the opportunity to learn how to analyze the raw data, and some expressed the desire to work on data from different industry settings (especially those who were not interested in transportation related jobs). Therefore, we decided to abandon these exercises in the second implementation. It is important to note that in both instances, students were allowed to ask questions to clarify the case information and request technical support during class time prior to project submission.

In terms of content, lectures on facility location were delivered in class (Murphy and Knemeyer 2018). During class time, students were asked to propose different ways to measure general factors influencing facility location, which helped them become familiar with certain selection criteria used in the case and build confidence when collecting publicly available data. A brief introduction to the multicriteria decision-making process with an example of the factor-rating method was also provided. Finally, the structure and idea behind optimization programs were explained to students, along with a walk-through demonstration of Open Solver (detailed instructions can be found in the teaching notes).

7.2. Case Adaptation

This case emulates the process of selecting a facility location through multiple layers of screening. The base case follows closely the basic concepts in facility location. Specifically, it focuses on (1) initiating the set of three-digit zip codes for consideration and (2) using distance-based approach (i.e., optimization) to finalize the most desirable zip code. The case extension, on the other hand, increases the subjectivity of the location solution by considering many criteria that are qualitative in nature. The presentation allows students to discuss and consider different perspectives adopted by their own team and others in the class.

With the previous structure, the case description and associated questions can be flexibly adjusted based on the instructor’s need, students’ background, and available time. First, this case can be taught without invoking optimization problems if the objective is to teach the general decision-making process. This setting shortens the preparation time for the case and accommodates the lack of technical skills in the student population. In this scenario, the instructor may want to develop more questions for guiding students in implementing specific evaluation methods (such as the factor-rating method). For example, when we first implemented the case, we did not ask guiding questions regarding the factor-rating method. Although most students reported their team efforts to prioritize different selection criteria and rank them to derive the final solution, only one team submitted their formal calculations of the total weighted scores in a summary table. In the second time, we alluded to an example of the method, more than half of them used a scoring scheme in their submission. Our experience indicates that if the instructor wishes to assess any method, it may be more effective to provide students with explicit questions regarding that method in the assignment.

Second, the case can be taught in a business analytics course to introduce optimization. The case description can be tailored to a more generic audience; for example, the logistics-career-related content could be excluded, and modeling-related questions could be added. Because the decision alternatives in our case are not massive, instructors can use the exercises as in Brusco (2022) to present an intuitive approach to solving discrete optimization problems.

Third, this case can be implemented without the presentation requirement. In fact, when we used the case the second time, we did not require a formal presentation from student teams for two reasons: (1) based on student feedback, teams appeared to spend substantial time discussing the problem to get to a consensus; and (2) we had enough in-class interactions with teams prior to the submission of the written reports. That said, some students expressed their desire to know the conclusion from other teams. Hence, we think it may be beneficial to students if the instructor could provide a short debrief after grading the project.

Finally, the case can be designed as an individual homework assignment with a follow-up in-class discussion. To increase engagement, the instructor could ask a couple of students to present their solutions while others will ask questions or challenge the presented solutions. The content can be scaled down to focus on selection criteria that are relevant to the provided lecture about facility location.