- Email: [email protected]
Improving the efficiency of a non-profit supply chain for the food insecure
Int. J. Production Economics 143 (2013) 248–255
Contents lists available at ScienceDirect
Int. J. Production Economics journal homepage: www.elsevier.com/locate/ijpe
Improving the efficiency of a non-profit supply chain for the food insecure S. Mohan a, M. Gopalakrishnan a,n, P.J. Mizzi b a b
Department of Supply Chain Management, W.P. Carey School of Business, Arizona State University, Tempe, AZ, United States Department of Economics, W.P. Carey School of Business, Arizona State University, Tempe, AZ, United States
a r t i c l e i n f o
a b s t r a c t
Available online 27 May 2011
This paper focuses on the operational planning issues in a non-profit supply chain that distributes food for the ‘‘food insecure’’. The study details how by partnering with a local University the organization was able to improve the efficiency and effectiveness of its operations that resulted in more food reaching the food insecure. A detailed simulation model of the warehouse operations where food is processed served as a framework for making changes that improved the efficiency of the operations in terms of handling extra volume without investing in additional warehouse space. In addition, proper demand planning, supply coordination and logistics integration were key drivers for improving the effectiveness of this food supply chain. & 2011 Elsevier B.V. All rights reserved.
Keywords: Humanitarian logistics Food insecurity Simulation Non-profit
1. Introduction Food insecurity is defined as the limited or uncertain availability of nutritionally adequate, safe foods, or a person’s inability to acquire personally acceptable foods in socially acceptable ways. The recession of 2008–2009 has exacerbated the food insecurity problem. For example, in the United States of America close to 49 million people were found to be ‘‘food insecure’’ up from 36 million in 1997 (DeParle, 2009; Nord et al., 2009). In Arizona, where our study was conducted, nearly 900,000 people experienced food insecurity in 2009 up by 85% over previous study estimates in 2006 (Hildebrand and Simpson, 2010). A shortage of supply is not the problem as a surplus of food exists, even in countries poorer than the United States. However, the challenge is to get the available food reach the needy. Various non-profit groups are working to provide the food to needy families. For example, in 2009, 37 million people obtained food through non-profit agencies in the United States (Wunderlich and Norwood, 2006; Hildebrand and Simpson, 2010). These organizations act as food intermediaries, managing the entire food supply chain for the ‘‘food insecure’’ by obtaining food from donor organizations and individuals, processing the food and distributing it to the hungry. The organizations are primarily run by volunteers to keep the cost low. However, their food supply chain is riddled with inefficiencies because of lack of knowledge in managing them. This study focuses on the operational issues in one such supply chain and describes how a simulation-based approach was used to improve
n
Corresponding author. E-mail addresses: [email protected] (S. Mohan), [email protected] (M. Gopalakrishnan), [email protected] (P.J. Mizzi). 0925-5273/$ - see front matter & 2011 Elsevier B.V. All rights reserved. doi:10.1016/j.ijpe.2011.05.019
the efficiency and productivity of a food reclamation center and also describes the strategies used to analyze the need and distribute food effectively. The goal of the supply chain is to minimize the operational costs and maximize the quantity of food reaching the food insecure. The rest of the paper is organized as follows. In the next section we discuss the relevant literature on humanitarian logistics. We describe the case of the non-profit supply chain, the challenges that it was facing to fulfill its obligation with a growing demand and the tactical and operational strategies that it deployed to successfully manage its growth crisis in the following sections. We conclude with directions for future research in the last section.
2. Literature review Humanitarian logistics represents adequate response to a disaster and/or humanitarian crisis. Managing food insecurity and making sure food reaches the needy is definitely a type of humanitarian logistics issue. Response includes, ‘‘preparedness, planning, assessment, appeal, mobilization, procurement, transportation, warehousing and distribution’’ (Ozbay and Ozguven, 2007). Humanitarian relief chains are generally, non-profit supply chains that coordinate assistance in the form of food, water, medicine, shelter and supplies to people affected by emergencies (Beamon and Balcik, 2008). Emergencies could vary from food insecurity caused primarily by economic conditions, to large-scale emergencies such as earthquakes or floods. Logistics is central to the coordination in humanitarian relief operations and also the most expensive part of the relief operation. Hence it needs to be managed in the most effective (‘‘doing the right things’’) and
S. Mohan et al. / Int. J. Production Economics 143 (2013) 248–255
efficient (‘‘doing things right’’) manner (Van Wassenhove, 2006; Gatignon et al., 2010). Oloruntoba and Gray (2006) state that coordination in humanitarian supply chains is also the ability to cope with uncertainties in the supply chain. One effective ‘‘coping’’ strategy is understanding the decoupling point in such supply chains where ‘‘downstream market-pull’’ (i.e., end customer demand) meets ‘‘upstream push’’ (forecast-based). The philosophy in humanitarian supply chains is that supplies are ‘‘pushed’’ to the crisis location in the immediate response phase, while the ‘‘pull’’ philosophy is applied in the reconstruction phase (Kovacs and Spens, 2007). Also, the upstream operations in the push phase would focus on improving the efficiencies of material (example food) flow and prepare ‘‘strategic inventory’’ that can be distributed through proper logistics to satisfy evolving needs of the end user (Oloruntoba and Gray, 2006). In our field work described later, a reclamation center becomes the ‘‘decoupling point’’ whose efficiency improvement is paramount to effective food distribution for the needy. Key success factors in managing humanitarian supply chains fall into two broad categories, namely, (a) preparedness and readiness and (b) unity of direction and cohesive control of responding agencies (Oloruntoba, 2010). Proper planning is an essential driver for good performance and our work talks about a simulation framework as a tool for food reclamation center planning. Lastly, effective performance measures would assist relief chain practitioners in their decisions and help improve effectiveness and efficiency of relief operations. This is turn increases the transparency and accountability of crisis response (Beamon and Balcik, 2008; Balcik et al., 2010). The non-profit humanitarian supply chain that we studied faced many of the same issues. In the next section, we describe the operations of the non-profit organization and its supply chain.
3. St. Vincent de Paul’s humanitarian supply chain The Society of Saint Vincent de Paul (SVdP) is an international organization present in 130 countries spanning five continents and counting 900,000 members. Founded in 1833, in Paris, France, by six University students desirous of helping the Parisian poor, the Society is one of the oldest charitable organizations. Since its inception, the mission of the Society is to serve the needs of the poor and to provide to those more fortunate, an opportunity to serve others. The Society has served homeless and low-income families in Arizona, USA since 1946 through a variety of programs including a free medical and dental clinic, dining rooms, transitional housing and an opportunity program to help people build job skills through volunteer service. Services are provided without regard to race, origin, religion or gender (Society of St. Vincent de Paul, Phoenix website, 2011). 3.1. Food supply chain The Society’s food reclamation center (FRC) collects nonperishable food items donated by grocery stores and community food drives. Most of the food comes from grocery stores that provide not saleable food and non-food items with damaged packaging or nearing their expiration date. FRC cleans, accounts, inspects, sorts and stores the food using employees and volunteers. It then assembles the food on pallets primarily for pick up by the Society’s Conferences of Charity. The conferences store the food in their pantries and distribute it free of charge to people seeking assistance. Normally, a conference serves people who live in the neighborhood and who have been referred to it. Conference volunteers also make home visits to establish need and deliver food. For example, in fiscal year 2009, roughly around 3000
249
volunteers in 73 conferences distributed around 400,000 food boxes to Arizona residents in need of food assistance. FRC also distributes food to the various dining rooms that use it for cooking and delivering meals to the people who visit them to dine on a daily basis (breakfast, lunch and dinner). A typical non-profit food bank in the United States is provided food through donations by individuals, grocers and other organizations. The food must be inspected for usability following strict standards to eliminate food packages that are damaged or having expired use-by dates. The food is then assembled in bulk lots by the food bank and distributed to other organizations that prepare the food through free meals or provide it in packaged form to the food insecure families. For the Society, the supply chain is more complex. The bulk of food is provided by a local grocery chain through regular deliveries. Additional food enters the system through food drives and other donations. All the food must be unloaded and then temporarily stored. The food from the grocer must be scanned to provide an accounting to the grocer to be used in their inventory management system and for tax benefit purposes. Once the food has been scanned, it can then be combined with other donated food and sorted in various food groups such as cereals, pastas, rice and so on. The food then goes through quality control before being assembled in pallets (‘‘strategic inventory’’, Oloruntoba and Gray, 2006) for pick-up by the Society’s conferences that then repackage the food for the food insecure families. Food obtained through the FRC also supports SVdP’s five dining rooms in the Phoenix metropolitan area. The Society provides hot meals to other non-profits in the area that do not have adequate kitchen facilities. Food not distributed through the conferences or used in the Society’s dining rooms is transferred to other food banks or distributed to the needy during Mass distribution events. Fig. 1 illustrates this food supply chain. 3.2. Supply chain challenges 3.2.1. Supply The supply is dependent on donations, food drives and delivery from grocery stores. Hence it is uncertain and constantly evolving. For example, the donation from the grocery chain has an organized process for collecting and reporting the donated food products (damaged, out-of-code and seasonal items—termed as shrinkage). However, the amount available for use varied from week to week. Furthermore, grocery chains continuously try to reduce their shrinkage rates (Grannis, 2010), which in turn affects the amount of food provided to the warehouse. Similarly, the donations from other sources are not predictable especially at times of economic uncertainty and have quite a variation in the quantity available for useful distribution and consumption. On the other hand there was consolidation in the grocery chains in Arizona around 2004 and hence the general volume of food available at the warehouse was estimated to increase tenfold (from one million pound per year to one million pound per month). Hence, an immediate challenge was to accommodate this increase in food coming in. The Board at SVdP was trying to decide whether to invest in a new warehouse or use the storage space in the current warehouse more effectively. The latter idea translates to increasing efficiencies in receiving, processing, storing and distributing food. 3.2.2. Demand The demand for food could be analyzed at two levels. At the macro level, the ‘‘need’’ was dependent on factors such as poverty, local economic conditions (unemployment rates), migration rates and cost of living. Hence, based on census data we could estimate
250
S. Mohan et al. / Int. J. Production Economics 143 (2013) 248–255
Supply
Other Food Banks
Operation
Grocery Stores
Food Drives and special events
Phoenix Food Reclamation Center Scanning
Tuscon Food Reclamation Center
Quality control Sorting Assembly
Distribution Dining Rooms
SVdP Conferences
Other Food Banks
Mass Distribution
Demand Needy
Fig. 1. Food supply chain for the ‘‘food insecure’’.
by zip code the poverty index and the general level of food insecurity. This insecurity stayed relatively stable in the short run, though migration trends and the state of the local economy could have an impact in the long run (Wunderlich and Norwood, 2006). However, the day-to-day demand addressed through the SVdP conferences was highly variable. This demand was a function of (a) people with food-need who showed-up at the conferences and soup kitchens and (b) the house visits made by SVdP personnel at the conferences responding to need. This demand is greatly affected by the local economy (Hildebrand and Simpson, 2010). Furthermore, the pantry space at the conferences that distributed food was limited and played a significant role in quantity of food distributed. In addition, employees from the conferences made a trip to the reclamation center where food was stored, at their convenience, to pick up food that could be carried in their personal vehicle. This limited the frequency and amount of food that could be taken back for distribution to the conferences which further exacerbated the ability to service the demand. Thus, even if food was available, it was often sitting in the reclamation center due to inefficiencies in the distribution process and storage constraints at the conferences. Hence it was clear that there was a mismatch between supply and demand resulting in two outcomes, both undesirable: (1) wastage of food due to shelf life and (2) unmet demand. Hence, it was critical to improve the efficiencies at the reclamation center and of the distribution processes to better respond to demand.
4. Research partnership and outcomes The Society approached the authors at the local state University to assist in addressing these challenges. The University formed a team of faculty and graduate students and worked with the Society and other
partners in the state (examples: other food banks, granting agencies and private sector) for a period of five years to deliver a combination of solutions. The faculty team was multi-disciplinary to tackle issues in economics, engineering and management. The student involvement happened at two levels, namely, graduate research assistants focused on specific projects and class-projects in graduate courses (Economics, Quantitative Business Analysis and Supply Chain Management courses), targeting specific areas that provided opportunities for data collection and analysis. Three areas were identified, namely: needs analysis, operational analysis and distribution analysis. The studies in these areas provided the comprehensive knowledge that has helped the Society handle its growth and uncertainties in an efficient and effective manner. The Society was actively involved in the studies, providing valuable time, information and feasibility check of ideas as they were developed. Also, the University–Society partnership helped the Society in obtaining grants to fund the structural and infrastructural needs. 4.1. Needs analysis This part of the study looked at economic models of poverty to establish the food-insecurity levels, which serve as an estimate of the demand. The federal governmental data released by state on food security levels and related predictors (in support of the National Nutrition Monitoring and Related Research Act of 1990 (NNMRR)) became the key input to this phase of the study. The most obvious characteristic for prediction of food security levels is household income but the federal government also reports noticeable higher food insecurity levels among certain high-risk populations, specifically, single-women headed households with children, hispanic households and black households. Based on this data our team developed an econometric model that also included the average income by census tract, the number of census tract residents over age 55 and the number of non-white
S. Mohan et al. / Int. J. Production Economics 143 (2013) 248–255
households in the census tract to predict food insecurity level in terms of number of households that are in need, established at a postal zip code level. The location of conferences with respect to these postal codes along with other independent food banks become the service points to service this need. The needs analysis provided a good estimate of the percentage of need that can be met and also the gap that needs to be addressed. 4.2. Operational analysis The focus of the operational analysis was at the food reclamation center level with the goal of increasing the throughput efficiency so that more food is available quicker for distribution. First we present the different steps at the reclamation center with their status at the starting of this project. Next, we describe the steps taken to improve efficiencies and their results. We provide details of the discrete event simulation model used to test and validate our recommendations for operational improvement. 4.2.1. Receive food The reclamation center processes start with trucks from the grocery chains and trucks picking food from other donors arriving at the receiving dock. At the beginning of the project there was only one receiving dock. Hence, if more than one truck arrived at the same time it led to congestion and waiting time for the truckers who were volunteering this service and had other commercial loads to haul. Also, at the beginning of the project, food was unloaded from the truck by one fork-lift truck that would place the unloaded pallet at any available space within the warehouse, so that the delivery truck could be released. Next, the pallets had to be weighed for accounting purposes. The same forklift truck would move the pallets of food from various locations to the weighing station, weigh the pallet and then move it to the scanning station. This resulted in excessive and unnecessary handling of the pallets and a hap hazard layout of the incoming pallets of food. 4.2.2. Scanning After the weighing operation, the pallets were unloaded in front of a single scanner. Here, items go through a manual sorting operation to separate food and non-food items. After the sorting operation, the items are scanned for accounting and tax purposes. The Society rendered this as a service to the grocery chains for their donations. 4.2.3. Quality control The sorted and scanned items were stored in cardboard banana boxes. The banana boxes were stored thirty-six to a pallet. The fork-lift truck them moved the pallets with the banana boxes to the quality control section. Here, three volunteers inspected the food items for expiration dates and any other aspects that would make the items unsafe for use. This stage was laborious, time-consuming and was staffed by volunteers. The food that
passed inspection is put back into the banana boxes and sent to the next stage for processing. At the start of our study, there were three quality control stations. 4.2.4. Sort, assemble and store This last step separated the inspected items into different categories (example: dry food like pasta versus canned vegetables) and packed them in boxes for storage. The boxes were stored in dedicated storage areas for pick up and distribution to the conferences. If the reclamation center did not send out food to the conferences and other venues frequently and in sufficient quantities the final storage also became congested. 4.3. Simulation model In order to test different options to improve the efficiency of the reclamation center, we had to first measure the performance of the existing system. We developed a detailed discrete event simulation model to mimic the reclamation center operations. The model was developed using Arena simulation software. We did extensive data collection and analysis on the following input parameters: a. Inbound food: we analyzed the detailed logs on truck arrivals at the loading docks that covered a two-year period. This analysis helped determining the arrival rates and the quantity of food (in pounds) received daily. b. Process times—unloading, scanning, quality control and assembly and storage: we collected data on these process times using detailed time studies. The data was collected by graduate students and volunteers at the Society. c. Layout and transportation distances within the warehouse: layout diagrams and actual measurement of distances that the food moves as it travels from unloading to assembly helped determine the flow patterns. Once we obtained the data, we fit statistical probability distributions to characterize the underlying the variability in the parameters. Example of one such analysis is shown in Table 1. This table shows the statistical distributions used to represent the pounds of food received everyday at the reclamation center. We developed similar probability distributions for the various process times. The percentage of goods eliminated as defects at each stage was also determined using the data collection study. The model was used to analyze the current situation to obtain metrics such as throughput time, utilizations and overtime. Next, various ‘‘what-if’’ analyses were conducted to study the impact of changes in improving the productivity and efficiency of the reclamation center. Fig. 2 shows the various what-if analyses we performed and the resulting models. The original model is represented as Model 1 in the figure. We explain the reasons for the several test scenarios below and identify the corresponding model in Fig. 2. Baker (2008) identifies excess docks, smaller pick
Table 1 Statistical Distribution Fit of incoming food (in pounds).
Monday Tuesday Wednesday Thursday Friday Saturday
251
Scanned input
Non-scanned input
TRIA( 0.001, 1.57e þ004, 3.13e þ 004) TRIA( 0.001, 1.47e þ004, 4.89e þ 004) 0.001 þEXPO(1.44e þ 004) 0.001 þEXPO(6.96e þ 003) TRIA(8.85eþ 003, 2.09e þ004, 3.29e þ004) 0
1.5e þ003 þ WEIB(3.38eþ003, 0.746) 1.56e þ003 þ EXPO(8.16e þ003) 150 þ GAMM(1.18eþ 004, 0.609) 525 þ WEIB(3.5eþ 003, 0.663) 323 þ 1.55eþ 004 BETA(0.408, 0.66) 0.001þ WEIB(1.31eþ 003, 0.813)
252
S. Mohan et al. / Int. J. Production Economics 143 (2013) 248–255
Fig. 2. What-if analysis and corresponding simulation models.
bins, flexible storage, staging and additional warehouse space as a few critical variables in improving the efficiency of distribution centers. At the reclamation center, many of these variables were identified as the true bottlenecks, hindering efficient movement of food through the reclamation center and leading to excessive overtime. The following improvements that include some of the variables identified by Baker (2008) were investigated using the simulation model. 4.3.1. Addition of a second dock During our initial study, one of the grocery chains acquired a smaller chain and added several stores to its operation. Hence, there was a dramatic increase in the quantity of food coming in from that chain. The scanning operation became a huge bottleneck. So, Model 2 studied the impact of adding a second loading dock and a second scanning station to process the incoming food quickly. The addition of the second dock was critical to handle the additional volume. Management also wanted to investigate the option of adding longer conveyors to feed the food to the scanners. We worked with the Society in identifying funding sources to finance the construction of the second dock. While this change reduced the percentage of time the scanners were busy, it also highlighted the fact that scanners were starving for work frequently since the pallets of food were located all over the warehouse and the fork-lift truck was unable to keep the scanners busy. Also, there was excessive movement of materials. 4.3.2. Addition of a dedicated staging area The results of Model 2 motivated the researchers and operations personnel to seriously consider designing a dedicated
staging area for incoming pallets. On working with the operations personnel, it was determined that it would be feasible to move a few items such as clothing and furniture that were using the warehouse space to other locations, tear down some non-load bearing walls and make room for a staging area close to the loading docks and the scanners. In order to study the impact of this move, we developed Model 3 with a dedicated staging area. Other changes the Society wanted to try were using a fork-lift truck with an integrated weighing machine and adding another fork-lift truck. The dedicated staging area minimized the movement of materials and streamlined the flow in the reclamation center. The elimination of additional material movement for weighing eased the flow some more. The addition of the second fork-lift truck avoided starving at the scanners and the scanned food quickly moved to the quality control stations.
4.3.3. Lengthening downstream conveyors and preliminary quality check Model 3 showed that with a dedicated staging area, two forklift trucks with integrated weighing scales and two scanners with longer conveyors, incoming food could be weighed, staged and scanned relatively quickly. However, this only moved the bottleneck downstream. The boxes of food arrived at the three quality control stations much faster and the volunteers at these stations could not keep pace. Eventually, this led to the scanners being blocked. In an effort to ease the workload on the quality control stations, management wanted to increase the length of the conveyors from the scanners to the boxing area in order to complete some of the initial quality control as the scanned goods move down the conveyor (Model 4).
S. Mohan et al. / Int. J. Production Economics 143 (2013) 248–255
The Society with assistance from the University was able to obtain funds to buy their own trucks. A routing schedule established the ‘‘milk runs’’ to be made on a daily basis. Fig. 5 compares the old and improved approaches. The Society also coordinated its efforts with other food banks to increase the amount of food distributed. Lastly, the Society conducted more mass distribution events. They staged events in high schools in the food insecure areas where thousands of pounds of food were disbursed in a single day. All of these ideas have led to more food reaching the food insecure.
4.3.4. Addition of a fourth quality control station A sample study helped determine that volunteers would be able to complete about 10% of the quality control work as the scanned good moved down the conveyor. The results from Model 4 showed that with the addition of two volunteers along the conveyor, the workload for the quality control stations was reduced by an average 11%, based on a simulation of an entire week. This, however, was not sufficient to relieve the heavy workload on the quality control stations. Hence the next model (Model 5) added another quality control station. The simulation model showed that the addition of a quality control station kept the work across the various stages more balanced than before. After trying the various scenarios and demonstrating the results using simulation-animation, management was ready to implement the changes. The conditions described by Model 5 (two unloading docks, dedicated staging area, two fork-lift trucks with integrated scales, two scanners with long conveyor feed-in lines, pre-sorting and preliminary quality control on the conveyor from scanner to boxing and four quality control stations) were implemented at the reclamation center. The outcomes from these changes were profound. The addition of the second dock and the changes to unloading, weighing and staging improved the unloading times and hence the trucks were released sooner leading to lesser congestion. The improvements made at scanning and quality control really improved throughput. Hence the FRC was able to handle the increase in volume without building a new warehouse. Figs. 3 and 4 show the initial and final layouts and how the process improvements resulted in an orderly flow and reduced congestion.
5. Conclusion A crisis or disaster can be the result of a wide variety of adverse conditions including economic conditions, natural disasters, environmental threats, financial meltdowns and surprise attacks. An overriding feature of crisis and disaster is the high degree of uncertainty in terms of supply and demand for aid, and the humanitarian logistics challenge is that of ‘‘getting the right aid to the right people, at the right time’’. The key elements of tackling this challenge include building on a correct inventory of needs, a rapid surge of goods, transportation of goods and a delivery system that is as efficient as the circumstances allow it to be (Boin et al., 2010). This paper discussed the logistics elements involved in distributing food for the food insecure. The reclamation center plays the key role of acquiring and assembling food for distribution. The use of a simulation model to analyze and improve the efficiency and productivity of this warehouse was detailed. Speed of response is a critical performance variable and the operational changes made in the warehouse helped improve the throughput of the reclamation center. In addition, the fulfillment of demand was also addressed through an effective distribution strategy. The result has been a combination of increased mouths fed and minimized wastage of food. Future directions include the following:
4.4. Distribution analysis The improvements at the reclamation center increased throughput within the warehouse by 138%. However, the bottleneck now shifted to distribution (i.e., moving food from warehouse through the conferences to the ‘‘food insecure’’). The prevailing approach of conferences picking up of food from the warehouse was inefficient. Also, when there is a food shortage at a conference, it was not immediately replenished. Hence, the proposed distribution model involved the Society making deliveries to vantage locations from where the local conferences pick up food (similar to cross-docking).
1. Analyzing the role of information technology in improving the needs analysis and coordination: the Society is currently trying to establish a partnership with a large technology company that could infuse information technology along the supply
Cold storage
1
1
Loading dock;
4 Quality control
4
3
2
Wood Shop
253
1st - Clothing 2nd - Storage
2 Fork lift
Mattresses
3 Scanner
Food pallet Fig. 3. Initial layout of the reclamation center.
254
S. Mohan et al. / Int. J. Production Economics 143 (2013) 248–255
4
Cold storage
3
2
1
1
Loading dock (capable of un-loading three trucks at a time)
2
Fork- lift truck with weighing scale Scanners (3) with pre-quality check and sorting as the items are scanned
3 4
Quality control section with increased capacity Fig. 4. Modified Food Reclamation center layout.
purpose of real time updating of inventory and scheduling of pickups and deliveries (Noe, 2007). 2. Developing direct distribution practices: discussions are underway to see if the sorting and scanning of donated food from the grocery chains can be done at the warehouses of the grocery chains (rather than the reclamation center of the Society) and distributed directly to the conferences through a cross-docking approach. This will further improve the turnaround time in food delivery and minimize Society’s operational costs. 3. Better coordination with other food banks and also carrying the best practices to other non-profit food reclamation centers: efforts are underway to form a consortium of companies in this humanitarian space and the University playing a critical role in coordinating the different activities at the local and even at an international level.
C2 C1 C3 C6
RC
C5
C4
RC: Reclamation center; C1-C6 Examples of six Conferences Approach in the push system: Conferences made trips to the RC to pick up food.
A C2 C1 C3
References
C6
RC
B
C C5
C4
RC: Reclamation center; C1-C6 Examples of six Conferences; A, B, C Examples of clusters. Approach in the push-pull system: Society made milk runs to vantage points in the clusters. Conferences cross-docked at these points to pick up food, which improved their responsiveness in delivering food for the need.
Fig. 5. Food Distribution Analysis.
chain (example: at the conferences and with the delivery trucks). The primary objective is to create a network that will connect distant computers and mobile platforms for the
Baker, P., 2008. The design and operations of distribution centers in Agile supply chains. International Journal of Production Economics 111, 27–41. Balcik, B., Beamon, B.M., Krejci, C.C., Muramatsu, K.M., Ramirez, M., 2010. Coordination in humanitarian relief chains: practices, challenges and opportunities. International Journal of Production Economics 126, 22–34. Beamon, B.M., Balcik, B., 2008. Performance measurement in humanitarian supply chains. International Journal of Public Sector Management 21 (1), 4–25. Boin, A., Kelle, P., Whybark, C.D., 2010. Resilient supply chains for extreme situations: outlining a new field of study. International Journal of Production Economics 126, 1–6. DeParle, J., 2009. Hunger in US at a 14-Year High. The New York Times. November 17, 2009. Gatignon, A., VanWassenhove, L.N., Charles, A., 2010. The Yogyakarta earthquake: humanitarian relief through IFRC’s decentralized supply chain. International Journal of Production Economics 126, 102–110. Grannis, K., 2010. Retail Fraud, Shoplifting Rates Decrease, According to National Retail Security Survey. /http://www.nrf.com/modules.php?name=News&op=viewlive &sp_id=945S (accessed 26.04.11.). Hildebrand, G., Simpson, B., 2010. Massive Hunger in America Study Shows More Arizonans Seeking Emergency Food Assistance. /www.azfoodbanks.orgS. February 2, 2010 (accessed 29.07.10.).
S. Mohan et al. / Int. J. Production Economics 143 (2013) 248–255
Kovacs, G., Spens, K.M., 2007. Humanitarian logistics in disaster relief operations. International Journal of Physical Distribution and Logistics Management 36 (2), 99–114. Noe, S., 2007. Network solutions for a non-profit reclamation center. In: Proceedings of 38th Annual Meeting—Decision Sciences Institute. Phoenix, Arizona. November 17–20, 2007. Nord, M., Andrew, M., Carlson, S., 2009. House Hold Food Security in the United States, 2008. United Stated Department of Agriculture. Economic Research Report #83. November 2009. Ozbay, K., Ozguven, E.E., 2007. A stochastic humanitarian inventory control model for disaster planning. In: Proceedings of 86th Annual Meeting. Transportation Research Board. Washington, D.C.
255
Oloruntoba, R., Gray, R., 2006. Humanitarian aid: an Agile supply chain. Supply Chain Management—An International Journal 11 (2), 115–120. Oloruntoba, R., 2010. A documentary analysis of the Cyclone Larry emergency relief chain: some key success factors. International journal of Production Economics 126 (1), 85–101. Society of St. Vincent de Paul, Phoenix website, 2011. /http://www.stvincentde paul.net/S (accessed 26.04.11.). Van Wassenhove, L.N., 2006. Blackett memorial lecture. Humanitarian aid logistics: supply chain management in high gear. Journal of the Operational Research Society 57, 475–489. Wunderlich, G.S., Norwood, J.L., 2006. Food insecurity and hunger in the United States—an assessment of the measure. /http://www.nap.edu/openbook. php?record_id=11578&page=R1S (accessed 26.04.11.).
Contents lists available at ScienceDirect
Int. J. Production Economics journal homepage: www.elsevier.com/locate/ijpe
Improving the efficiency of a non-profit supply chain for the food insecure S. Mohan a, M. Gopalakrishnan a,n, P.J. Mizzi b a b
Department of Supply Chain Management, W.P. Carey School of Business, Arizona State University, Tempe, AZ, United States Department of Economics, W.P. Carey School of Business, Arizona State University, Tempe, AZ, United States
a r t i c l e i n f o
a b s t r a c t
Available online 27 May 2011
This paper focuses on the operational planning issues in a non-profit supply chain that distributes food for the ‘‘food insecure’’. The study details how by partnering with a local University the organization was able to improve the efficiency and effectiveness of its operations that resulted in more food reaching the food insecure. A detailed simulation model of the warehouse operations where food is processed served as a framework for making changes that improved the efficiency of the operations in terms of handling extra volume without investing in additional warehouse space. In addition, proper demand planning, supply coordination and logistics integration were key drivers for improving the effectiveness of this food supply chain. & 2011 Elsevier B.V. All rights reserved.
Keywords: Humanitarian logistics Food insecurity Simulation Non-profit
1. Introduction Food insecurity is defined as the limited or uncertain availability of nutritionally adequate, safe foods, or a person’s inability to acquire personally acceptable foods in socially acceptable ways. The recession of 2008–2009 has exacerbated the food insecurity problem. For example, in the United States of America close to 49 million people were found to be ‘‘food insecure’’ up from 36 million in 1997 (DeParle, 2009; Nord et al., 2009). In Arizona, where our study was conducted, nearly 900,000 people experienced food insecurity in 2009 up by 85% over previous study estimates in 2006 (Hildebrand and Simpson, 2010). A shortage of supply is not the problem as a surplus of food exists, even in countries poorer than the United States. However, the challenge is to get the available food reach the needy. Various non-profit groups are working to provide the food to needy families. For example, in 2009, 37 million people obtained food through non-profit agencies in the United States (Wunderlich and Norwood, 2006; Hildebrand and Simpson, 2010). These organizations act as food intermediaries, managing the entire food supply chain for the ‘‘food insecure’’ by obtaining food from donor organizations and individuals, processing the food and distributing it to the hungry. The organizations are primarily run by volunteers to keep the cost low. However, their food supply chain is riddled with inefficiencies because of lack of knowledge in managing them. This study focuses on the operational issues in one such supply chain and describes how a simulation-based approach was used to improve
n
Corresponding author. E-mail addresses: [email protected] (S. Mohan), [email protected] (M. Gopalakrishnan), [email protected] (P.J. Mizzi). 0925-5273/$ - see front matter & 2011 Elsevier B.V. All rights reserved. doi:10.1016/j.ijpe.2011.05.019
the efficiency and productivity of a food reclamation center and also describes the strategies used to analyze the need and distribute food effectively. The goal of the supply chain is to minimize the operational costs and maximize the quantity of food reaching the food insecure. The rest of the paper is organized as follows. In the next section we discuss the relevant literature on humanitarian logistics. We describe the case of the non-profit supply chain, the challenges that it was facing to fulfill its obligation with a growing demand and the tactical and operational strategies that it deployed to successfully manage its growth crisis in the following sections. We conclude with directions for future research in the last section.
2. Literature review Humanitarian logistics represents adequate response to a disaster and/or humanitarian crisis. Managing food insecurity and making sure food reaches the needy is definitely a type of humanitarian logistics issue. Response includes, ‘‘preparedness, planning, assessment, appeal, mobilization, procurement, transportation, warehousing and distribution’’ (Ozbay and Ozguven, 2007). Humanitarian relief chains are generally, non-profit supply chains that coordinate assistance in the form of food, water, medicine, shelter and supplies to people affected by emergencies (Beamon and Balcik, 2008). Emergencies could vary from food insecurity caused primarily by economic conditions, to large-scale emergencies such as earthquakes or floods. Logistics is central to the coordination in humanitarian relief operations and also the most expensive part of the relief operation. Hence it needs to be managed in the most effective (‘‘doing the right things’’) and
S. Mohan et al. / Int. J. Production Economics 143 (2013) 248–255
efficient (‘‘doing things right’’) manner (Van Wassenhove, 2006; Gatignon et al., 2010). Oloruntoba and Gray (2006) state that coordination in humanitarian supply chains is also the ability to cope with uncertainties in the supply chain. One effective ‘‘coping’’ strategy is understanding the decoupling point in such supply chains where ‘‘downstream market-pull’’ (i.e., end customer demand) meets ‘‘upstream push’’ (forecast-based). The philosophy in humanitarian supply chains is that supplies are ‘‘pushed’’ to the crisis location in the immediate response phase, while the ‘‘pull’’ philosophy is applied in the reconstruction phase (Kovacs and Spens, 2007). Also, the upstream operations in the push phase would focus on improving the efficiencies of material (example food) flow and prepare ‘‘strategic inventory’’ that can be distributed through proper logistics to satisfy evolving needs of the end user (Oloruntoba and Gray, 2006). In our field work described later, a reclamation center becomes the ‘‘decoupling point’’ whose efficiency improvement is paramount to effective food distribution for the needy. Key success factors in managing humanitarian supply chains fall into two broad categories, namely, (a) preparedness and readiness and (b) unity of direction and cohesive control of responding agencies (Oloruntoba, 2010). Proper planning is an essential driver for good performance and our work talks about a simulation framework as a tool for food reclamation center planning. Lastly, effective performance measures would assist relief chain practitioners in their decisions and help improve effectiveness and efficiency of relief operations. This is turn increases the transparency and accountability of crisis response (Beamon and Balcik, 2008; Balcik et al., 2010). The non-profit humanitarian supply chain that we studied faced many of the same issues. In the next section, we describe the operations of the non-profit organization and its supply chain.
3. St. Vincent de Paul’s humanitarian supply chain The Society of Saint Vincent de Paul (SVdP) is an international organization present in 130 countries spanning five continents and counting 900,000 members. Founded in 1833, in Paris, France, by six University students desirous of helping the Parisian poor, the Society is one of the oldest charitable organizations. Since its inception, the mission of the Society is to serve the needs of the poor and to provide to those more fortunate, an opportunity to serve others. The Society has served homeless and low-income families in Arizona, USA since 1946 through a variety of programs including a free medical and dental clinic, dining rooms, transitional housing and an opportunity program to help people build job skills through volunteer service. Services are provided without regard to race, origin, religion or gender (Society of St. Vincent de Paul, Phoenix website, 2011). 3.1. Food supply chain The Society’s food reclamation center (FRC) collects nonperishable food items donated by grocery stores and community food drives. Most of the food comes from grocery stores that provide not saleable food and non-food items with damaged packaging or nearing their expiration date. FRC cleans, accounts, inspects, sorts and stores the food using employees and volunteers. It then assembles the food on pallets primarily for pick up by the Society’s Conferences of Charity. The conferences store the food in their pantries and distribute it free of charge to people seeking assistance. Normally, a conference serves people who live in the neighborhood and who have been referred to it. Conference volunteers also make home visits to establish need and deliver food. For example, in fiscal year 2009, roughly around 3000
249
volunteers in 73 conferences distributed around 400,000 food boxes to Arizona residents in need of food assistance. FRC also distributes food to the various dining rooms that use it for cooking and delivering meals to the people who visit them to dine on a daily basis (breakfast, lunch and dinner). A typical non-profit food bank in the United States is provided food through donations by individuals, grocers and other organizations. The food must be inspected for usability following strict standards to eliminate food packages that are damaged or having expired use-by dates. The food is then assembled in bulk lots by the food bank and distributed to other organizations that prepare the food through free meals or provide it in packaged form to the food insecure families. For the Society, the supply chain is more complex. The bulk of food is provided by a local grocery chain through regular deliveries. Additional food enters the system through food drives and other donations. All the food must be unloaded and then temporarily stored. The food from the grocer must be scanned to provide an accounting to the grocer to be used in their inventory management system and for tax benefit purposes. Once the food has been scanned, it can then be combined with other donated food and sorted in various food groups such as cereals, pastas, rice and so on. The food then goes through quality control before being assembled in pallets (‘‘strategic inventory’’, Oloruntoba and Gray, 2006) for pick-up by the Society’s conferences that then repackage the food for the food insecure families. Food obtained through the FRC also supports SVdP’s five dining rooms in the Phoenix metropolitan area. The Society provides hot meals to other non-profits in the area that do not have adequate kitchen facilities. Food not distributed through the conferences or used in the Society’s dining rooms is transferred to other food banks or distributed to the needy during Mass distribution events. Fig. 1 illustrates this food supply chain. 3.2. Supply chain challenges 3.2.1. Supply The supply is dependent on donations, food drives and delivery from grocery stores. Hence it is uncertain and constantly evolving. For example, the donation from the grocery chain has an organized process for collecting and reporting the donated food products (damaged, out-of-code and seasonal items—termed as shrinkage). However, the amount available for use varied from week to week. Furthermore, grocery chains continuously try to reduce their shrinkage rates (Grannis, 2010), which in turn affects the amount of food provided to the warehouse. Similarly, the donations from other sources are not predictable especially at times of economic uncertainty and have quite a variation in the quantity available for useful distribution and consumption. On the other hand there was consolidation in the grocery chains in Arizona around 2004 and hence the general volume of food available at the warehouse was estimated to increase tenfold (from one million pound per year to one million pound per month). Hence, an immediate challenge was to accommodate this increase in food coming in. The Board at SVdP was trying to decide whether to invest in a new warehouse or use the storage space in the current warehouse more effectively. The latter idea translates to increasing efficiencies in receiving, processing, storing and distributing food. 3.2.2. Demand The demand for food could be analyzed at two levels. At the macro level, the ‘‘need’’ was dependent on factors such as poverty, local economic conditions (unemployment rates), migration rates and cost of living. Hence, based on census data we could estimate
250
S. Mohan et al. / Int. J. Production Economics 143 (2013) 248–255
Supply
Other Food Banks
Operation
Grocery Stores
Food Drives and special events
Phoenix Food Reclamation Center Scanning
Tuscon Food Reclamation Center
Quality control Sorting Assembly
Distribution Dining Rooms
SVdP Conferences
Other Food Banks
Mass Distribution
Demand Needy
Fig. 1. Food supply chain for the ‘‘food insecure’’.
by zip code the poverty index and the general level of food insecurity. This insecurity stayed relatively stable in the short run, though migration trends and the state of the local economy could have an impact in the long run (Wunderlich and Norwood, 2006). However, the day-to-day demand addressed through the SVdP conferences was highly variable. This demand was a function of (a) people with food-need who showed-up at the conferences and soup kitchens and (b) the house visits made by SVdP personnel at the conferences responding to need. This demand is greatly affected by the local economy (Hildebrand and Simpson, 2010). Furthermore, the pantry space at the conferences that distributed food was limited and played a significant role in quantity of food distributed. In addition, employees from the conferences made a trip to the reclamation center where food was stored, at their convenience, to pick up food that could be carried in their personal vehicle. This limited the frequency and amount of food that could be taken back for distribution to the conferences which further exacerbated the ability to service the demand. Thus, even if food was available, it was often sitting in the reclamation center due to inefficiencies in the distribution process and storage constraints at the conferences. Hence it was clear that there was a mismatch between supply and demand resulting in two outcomes, both undesirable: (1) wastage of food due to shelf life and (2) unmet demand. Hence, it was critical to improve the efficiencies at the reclamation center and of the distribution processes to better respond to demand.
4. Research partnership and outcomes The Society approached the authors at the local state University to assist in addressing these challenges. The University formed a team of faculty and graduate students and worked with the Society and other
partners in the state (examples: other food banks, granting agencies and private sector) for a period of five years to deliver a combination of solutions. The faculty team was multi-disciplinary to tackle issues in economics, engineering and management. The student involvement happened at two levels, namely, graduate research assistants focused on specific projects and class-projects in graduate courses (Economics, Quantitative Business Analysis and Supply Chain Management courses), targeting specific areas that provided opportunities for data collection and analysis. Three areas were identified, namely: needs analysis, operational analysis and distribution analysis. The studies in these areas provided the comprehensive knowledge that has helped the Society handle its growth and uncertainties in an efficient and effective manner. The Society was actively involved in the studies, providing valuable time, information and feasibility check of ideas as they were developed. Also, the University–Society partnership helped the Society in obtaining grants to fund the structural and infrastructural needs. 4.1. Needs analysis This part of the study looked at economic models of poverty to establish the food-insecurity levels, which serve as an estimate of the demand. The federal governmental data released by state on food security levels and related predictors (in support of the National Nutrition Monitoring and Related Research Act of 1990 (NNMRR)) became the key input to this phase of the study. The most obvious characteristic for prediction of food security levels is household income but the federal government also reports noticeable higher food insecurity levels among certain high-risk populations, specifically, single-women headed households with children, hispanic households and black households. Based on this data our team developed an econometric model that also included the average income by census tract, the number of census tract residents over age 55 and the number of non-white
S. Mohan et al. / Int. J. Production Economics 143 (2013) 248–255
households in the census tract to predict food insecurity level in terms of number of households that are in need, established at a postal zip code level. The location of conferences with respect to these postal codes along with other independent food banks become the service points to service this need. The needs analysis provided a good estimate of the percentage of need that can be met and also the gap that needs to be addressed. 4.2. Operational analysis The focus of the operational analysis was at the food reclamation center level with the goal of increasing the throughput efficiency so that more food is available quicker for distribution. First we present the different steps at the reclamation center with their status at the starting of this project. Next, we describe the steps taken to improve efficiencies and their results. We provide details of the discrete event simulation model used to test and validate our recommendations for operational improvement. 4.2.1. Receive food The reclamation center processes start with trucks from the grocery chains and trucks picking food from other donors arriving at the receiving dock. At the beginning of the project there was only one receiving dock. Hence, if more than one truck arrived at the same time it led to congestion and waiting time for the truckers who were volunteering this service and had other commercial loads to haul. Also, at the beginning of the project, food was unloaded from the truck by one fork-lift truck that would place the unloaded pallet at any available space within the warehouse, so that the delivery truck could be released. Next, the pallets had to be weighed for accounting purposes. The same forklift truck would move the pallets of food from various locations to the weighing station, weigh the pallet and then move it to the scanning station. This resulted in excessive and unnecessary handling of the pallets and a hap hazard layout of the incoming pallets of food. 4.2.2. Scanning After the weighing operation, the pallets were unloaded in front of a single scanner. Here, items go through a manual sorting operation to separate food and non-food items. After the sorting operation, the items are scanned for accounting and tax purposes. The Society rendered this as a service to the grocery chains for their donations. 4.2.3. Quality control The sorted and scanned items were stored in cardboard banana boxes. The banana boxes were stored thirty-six to a pallet. The fork-lift truck them moved the pallets with the banana boxes to the quality control section. Here, three volunteers inspected the food items for expiration dates and any other aspects that would make the items unsafe for use. This stage was laborious, time-consuming and was staffed by volunteers. The food that
passed inspection is put back into the banana boxes and sent to the next stage for processing. At the start of our study, there were three quality control stations. 4.2.4. Sort, assemble and store This last step separated the inspected items into different categories (example: dry food like pasta versus canned vegetables) and packed them in boxes for storage. The boxes were stored in dedicated storage areas for pick up and distribution to the conferences. If the reclamation center did not send out food to the conferences and other venues frequently and in sufficient quantities the final storage also became congested. 4.3. Simulation model In order to test different options to improve the efficiency of the reclamation center, we had to first measure the performance of the existing system. We developed a detailed discrete event simulation model to mimic the reclamation center operations. The model was developed using Arena simulation software. We did extensive data collection and analysis on the following input parameters: a. Inbound food: we analyzed the detailed logs on truck arrivals at the loading docks that covered a two-year period. This analysis helped determining the arrival rates and the quantity of food (in pounds) received daily. b. Process times—unloading, scanning, quality control and assembly and storage: we collected data on these process times using detailed time studies. The data was collected by graduate students and volunteers at the Society. c. Layout and transportation distances within the warehouse: layout diagrams and actual measurement of distances that the food moves as it travels from unloading to assembly helped determine the flow patterns. Once we obtained the data, we fit statistical probability distributions to characterize the underlying the variability in the parameters. Example of one such analysis is shown in Table 1. This table shows the statistical distributions used to represent the pounds of food received everyday at the reclamation center. We developed similar probability distributions for the various process times. The percentage of goods eliminated as defects at each stage was also determined using the data collection study. The model was used to analyze the current situation to obtain metrics such as throughput time, utilizations and overtime. Next, various ‘‘what-if’’ analyses were conducted to study the impact of changes in improving the productivity and efficiency of the reclamation center. Fig. 2 shows the various what-if analyses we performed and the resulting models. The original model is represented as Model 1 in the figure. We explain the reasons for the several test scenarios below and identify the corresponding model in Fig. 2. Baker (2008) identifies excess docks, smaller pick
Table 1 Statistical Distribution Fit of incoming food (in pounds).
Monday Tuesday Wednesday Thursday Friday Saturday
251
Scanned input
Non-scanned input
TRIA( 0.001, 1.57e þ004, 3.13e þ 004) TRIA( 0.001, 1.47e þ004, 4.89e þ 004) 0.001 þEXPO(1.44e þ 004) 0.001 þEXPO(6.96e þ 003) TRIA(8.85eþ 003, 2.09e þ004, 3.29e þ004) 0
1.5e þ003 þ WEIB(3.38eþ003, 0.746) 1.56e þ003 þ EXPO(8.16e þ003) 150 þ GAMM(1.18eþ 004, 0.609) 525 þ WEIB(3.5eþ 003, 0.663) 323 þ 1.55eþ 004 BETA(0.408, 0.66) 0.001þ WEIB(1.31eþ 003, 0.813)
252
S. Mohan et al. / Int. J. Production Economics 143 (2013) 248–255
Fig. 2. What-if analysis and corresponding simulation models.
bins, flexible storage, staging and additional warehouse space as a few critical variables in improving the efficiency of distribution centers. At the reclamation center, many of these variables were identified as the true bottlenecks, hindering efficient movement of food through the reclamation center and leading to excessive overtime. The following improvements that include some of the variables identified by Baker (2008) were investigated using the simulation model. 4.3.1. Addition of a second dock During our initial study, one of the grocery chains acquired a smaller chain and added several stores to its operation. Hence, there was a dramatic increase in the quantity of food coming in from that chain. The scanning operation became a huge bottleneck. So, Model 2 studied the impact of adding a second loading dock and a second scanning station to process the incoming food quickly. The addition of the second dock was critical to handle the additional volume. Management also wanted to investigate the option of adding longer conveyors to feed the food to the scanners. We worked with the Society in identifying funding sources to finance the construction of the second dock. While this change reduced the percentage of time the scanners were busy, it also highlighted the fact that scanners were starving for work frequently since the pallets of food were located all over the warehouse and the fork-lift truck was unable to keep the scanners busy. Also, there was excessive movement of materials. 4.3.2. Addition of a dedicated staging area The results of Model 2 motivated the researchers and operations personnel to seriously consider designing a dedicated
staging area for incoming pallets. On working with the operations personnel, it was determined that it would be feasible to move a few items such as clothing and furniture that were using the warehouse space to other locations, tear down some non-load bearing walls and make room for a staging area close to the loading docks and the scanners. In order to study the impact of this move, we developed Model 3 with a dedicated staging area. Other changes the Society wanted to try were using a fork-lift truck with an integrated weighing machine and adding another fork-lift truck. The dedicated staging area minimized the movement of materials and streamlined the flow in the reclamation center. The elimination of additional material movement for weighing eased the flow some more. The addition of the second fork-lift truck avoided starving at the scanners and the scanned food quickly moved to the quality control stations.
4.3.3. Lengthening downstream conveyors and preliminary quality check Model 3 showed that with a dedicated staging area, two forklift trucks with integrated weighing scales and two scanners with longer conveyors, incoming food could be weighed, staged and scanned relatively quickly. However, this only moved the bottleneck downstream. The boxes of food arrived at the three quality control stations much faster and the volunteers at these stations could not keep pace. Eventually, this led to the scanners being blocked. In an effort to ease the workload on the quality control stations, management wanted to increase the length of the conveyors from the scanners to the boxing area in order to complete some of the initial quality control as the scanned goods move down the conveyor (Model 4).
S. Mohan et al. / Int. J. Production Economics 143 (2013) 248–255
The Society with assistance from the University was able to obtain funds to buy their own trucks. A routing schedule established the ‘‘milk runs’’ to be made on a daily basis. Fig. 5 compares the old and improved approaches. The Society also coordinated its efforts with other food banks to increase the amount of food distributed. Lastly, the Society conducted more mass distribution events. They staged events in high schools in the food insecure areas where thousands of pounds of food were disbursed in a single day. All of these ideas have led to more food reaching the food insecure.
4.3.4. Addition of a fourth quality control station A sample study helped determine that volunteers would be able to complete about 10% of the quality control work as the scanned good moved down the conveyor. The results from Model 4 showed that with the addition of two volunteers along the conveyor, the workload for the quality control stations was reduced by an average 11%, based on a simulation of an entire week. This, however, was not sufficient to relieve the heavy workload on the quality control stations. Hence the next model (Model 5) added another quality control station. The simulation model showed that the addition of a quality control station kept the work across the various stages more balanced than before. After trying the various scenarios and demonstrating the results using simulation-animation, management was ready to implement the changes. The conditions described by Model 5 (two unloading docks, dedicated staging area, two fork-lift trucks with integrated scales, two scanners with long conveyor feed-in lines, pre-sorting and preliminary quality control on the conveyor from scanner to boxing and four quality control stations) were implemented at the reclamation center. The outcomes from these changes were profound. The addition of the second dock and the changes to unloading, weighing and staging improved the unloading times and hence the trucks were released sooner leading to lesser congestion. The improvements made at scanning and quality control really improved throughput. Hence the FRC was able to handle the increase in volume without building a new warehouse. Figs. 3 and 4 show the initial and final layouts and how the process improvements resulted in an orderly flow and reduced congestion.
5. Conclusion A crisis or disaster can be the result of a wide variety of adverse conditions including economic conditions, natural disasters, environmental threats, financial meltdowns and surprise attacks. An overriding feature of crisis and disaster is the high degree of uncertainty in terms of supply and demand for aid, and the humanitarian logistics challenge is that of ‘‘getting the right aid to the right people, at the right time’’. The key elements of tackling this challenge include building on a correct inventory of needs, a rapid surge of goods, transportation of goods and a delivery system that is as efficient as the circumstances allow it to be (Boin et al., 2010). This paper discussed the logistics elements involved in distributing food for the food insecure. The reclamation center plays the key role of acquiring and assembling food for distribution. The use of a simulation model to analyze and improve the efficiency and productivity of this warehouse was detailed. Speed of response is a critical performance variable and the operational changes made in the warehouse helped improve the throughput of the reclamation center. In addition, the fulfillment of demand was also addressed through an effective distribution strategy. The result has been a combination of increased mouths fed and minimized wastage of food. Future directions include the following:
4.4. Distribution analysis The improvements at the reclamation center increased throughput within the warehouse by 138%. However, the bottleneck now shifted to distribution (i.e., moving food from warehouse through the conferences to the ‘‘food insecure’’). The prevailing approach of conferences picking up of food from the warehouse was inefficient. Also, when there is a food shortage at a conference, it was not immediately replenished. Hence, the proposed distribution model involved the Society making deliveries to vantage locations from where the local conferences pick up food (similar to cross-docking).
1. Analyzing the role of information technology in improving the needs analysis and coordination: the Society is currently trying to establish a partnership with a large technology company that could infuse information technology along the supply
Cold storage
1
1
Loading dock;
4 Quality control
4
3
2
Wood Shop
253
1st - Clothing 2nd - Storage
2 Fork lift
Mattresses
3 Scanner
Food pallet Fig. 3. Initial layout of the reclamation center.
254
S. Mohan et al. / Int. J. Production Economics 143 (2013) 248–255
4
Cold storage
3
2
1
1
Loading dock (capable of un-loading three trucks at a time)
2
Fork- lift truck with weighing scale Scanners (3) with pre-quality check and sorting as the items are scanned
3 4
Quality control section with increased capacity Fig. 4. Modified Food Reclamation center layout.
purpose of real time updating of inventory and scheduling of pickups and deliveries (Noe, 2007). 2. Developing direct distribution practices: discussions are underway to see if the sorting and scanning of donated food from the grocery chains can be done at the warehouses of the grocery chains (rather than the reclamation center of the Society) and distributed directly to the conferences through a cross-docking approach. This will further improve the turnaround time in food delivery and minimize Society’s operational costs. 3. Better coordination with other food banks and also carrying the best practices to other non-profit food reclamation centers: efforts are underway to form a consortium of companies in this humanitarian space and the University playing a critical role in coordinating the different activities at the local and even at an international level.
C2 C1 C3 C6
RC
C5
C4
RC: Reclamation center; C1-C6 Examples of six Conferences Approach in the push system: Conferences made trips to the RC to pick up food.
A C2 C1 C3
References
C6
RC
B
C C5
C4
RC: Reclamation center; C1-C6 Examples of six Conferences; A, B, C Examples of clusters. Approach in the push-pull system: Society made milk runs to vantage points in the clusters. Conferences cross-docked at these points to pick up food, which improved their responsiveness in delivering food for the need.
Fig. 5. Food Distribution Analysis.
chain (example: at the conferences and with the delivery trucks). The primary objective is to create a network that will connect distant computers and mobile platforms for the
Baker, P., 2008. The design and operations of distribution centers in Agile supply chains. International Journal of Production Economics 111, 27–41. Balcik, B., Beamon, B.M., Krejci, C.C., Muramatsu, K.M., Ramirez, M., 2010. Coordination in humanitarian relief chains: practices, challenges and opportunities. International Journal of Production Economics 126, 22–34. Beamon, B.M., Balcik, B., 2008. Performance measurement in humanitarian supply chains. International Journal of Public Sector Management 21 (1), 4–25. Boin, A., Kelle, P., Whybark, C.D., 2010. Resilient supply chains for extreme situations: outlining a new field of study. International Journal of Production Economics 126, 1–6. DeParle, J., 2009. Hunger in US at a 14-Year High. The New York Times. November 17, 2009. Gatignon, A., VanWassenhove, L.N., Charles, A., 2010. The Yogyakarta earthquake: humanitarian relief through IFRC’s decentralized supply chain. International Journal of Production Economics 126, 102–110. Grannis, K., 2010. Retail Fraud, Shoplifting Rates Decrease, According to National Retail Security Survey. /http://www.nrf.com/modules.php?name=News&op=viewlive &sp_id=945S (accessed 26.04.11.). Hildebrand, G., Simpson, B., 2010. Massive Hunger in America Study Shows More Arizonans Seeking Emergency Food Assistance. /www.azfoodbanks.orgS. February 2, 2010 (accessed 29.07.10.).
S. Mohan et al. / Int. J. Production Economics 143 (2013) 248–255
Kovacs, G., Spens, K.M., 2007. Humanitarian logistics in disaster relief operations. International Journal of Physical Distribution and Logistics Management 36 (2), 99–114. Noe, S., 2007. Network solutions for a non-profit reclamation center. In: Proceedings of 38th Annual Meeting—Decision Sciences Institute. Phoenix, Arizona. November 17–20, 2007. Nord, M., Andrew, M., Carlson, S., 2009. House Hold Food Security in the United States, 2008. United Stated Department of Agriculture. Economic Research Report #83. November 2009. Ozbay, K., Ozguven, E.E., 2007. A stochastic humanitarian inventory control model for disaster planning. In: Proceedings of 86th Annual Meeting. Transportation Research Board. Washington, D.C.
255
Oloruntoba, R., Gray, R., 2006. Humanitarian aid: an Agile supply chain. Supply Chain Management—An International Journal 11 (2), 115–120. Oloruntoba, R., 2010. A documentary analysis of the Cyclone Larry emergency relief chain: some key success factors. International journal of Production Economics 126 (1), 85–101. Society of St. Vincent de Paul, Phoenix website, 2011. /http://www.stvincentde paul.net/S (accessed 26.04.11.). Van Wassenhove, L.N., 2006. Blackett memorial lecture. Humanitarian aid logistics: supply chain management in high gear. Journal of the Operational Research Society 57, 475–489. Wunderlich, G.S., Norwood, J.L., 2006. Food insecurity and hunger in the United States—an assessment of the measure. /http://www.nap.edu/openbook. php?record_id=11578&page=R1S (accessed 26.04.11.).