Content uploaded by Adam Widera
Author content
All content in this area was uploaded by Adam Widera on Sep 16, 2021
Content may be subject to copyright.
Simulation in Produktion
und Logistik 2021
Jörg Franke & Peter Schuderer (Hrsg.)
Cuvillier Verlag, Göttingen 2021
Improving Blood Supply Chain Crisis Management
by Simulation-based Optimization
Dennis Horstkemper, WWU Münster, Münster (Germany), horstkemper@ercis.org
Melanie Reuter-Oppermann, TU Darmstadt, Darmstadt (Germany),
oppermann@is.tu-darmstadt.de
Michael Middelhoff, WWU Münster, Münster (Germany), middelhoff@ercis.org
Adam Widera, WWU Münster, Münster (Germany), widera@ercis.org
Bernd Hellingrath, WWU Münster, Münster (Germany), hellingrath@ercis.org
Abstract: Blood supply chains are a specialized type of supply chain dealing with
several legally independent actors that cooperate to facilitate the demand of a resource
with limited availability and shelf life. In this setting, adequate decision support tools
are needed to ensure good design choi1ces are made, as any interruption in the blood
supply chain endangers human lives. To handle the complexity of such a supply chain,
we propose a simulation-based optimization framework that enables the depiction of
strategic and operational decision-making throughout the overall BSC. We illustrate
the toolkits capabilities by applying it on a use case with the aim of improving the
utilization of mobile blood donation facilities during a crisis.
1 Introduction
Blood products are a very critical resource in healthcare systems, affecting several
procedures, such as scheduled or unplanned surgeries. Providing timely, sufficient
supply of blood products to match demand is a challenging task, reliant on volunteer
blood donations. Often, it is already challenging to fulfil everyday demands. During
a crisis, existing structures and resources might not be available anymore, while high
demand and supply fluctuations require short-term contingency plans. The COVID-19
pandemic has not only led to a shortage of blood products worldwide, but the loss of
crucial staff due to sickness or public health restrictions significantly affected the
blood supply chains (BSC) (Stanworth et al., 2020). Although decision support
systems or crisis management plans can support BSC management, several
characteristics of BSCs might explain the rather low uptake of current practices and
research from the areas of supply chain management and logistics. Firstly, the multi-
stakeholder setting of BSCs hinders the application of integrated supply chain design
and planning approaches. Secondly, the degree of uncertainty in demand and supply
178 Horstkemper, Dennis et al.
is significantly higher compared to other supply chains. Thirdly, it is not only the
diverse shelf life of blood products making the application domain highly dynamic in
terms of supply and demand: the need and stock of blood products is highly
interrelated with disruptions in other areas, such as power outages, mass-casualty
incidents, or large-scale disasters. Finally, a high degree of complexity and
interdependency can be identified when considering the different variations of blood
demands vs. supply as well as geographical and cultural contexts.
To handle the described challenges, we suggest the development and application of
combined simulation-optimization approaches. The main rationale lies in the ability
to vary and analyze assumptions that must be made for an optimization model, as for
example input values, relevant scenarios, and functional areas, within the approach.
Furthermore, stochastic influences can be tested within the simulation while keeping
the optimization model deterministic, reducing the mathematical complexity. Finally,
this approach combines rigorous methods with a continuous involvement of domain
experts. Simulation-based optimization approaches have been applied to BSCs and
production planning in disaster management before, yet existing approaches either
assume a centralized decision problem, only focus on smaller excerpts of the BSC, or
neglect the implications of disasters on the demand and supply of blood products
(Osorio et al., 2015; Ramezanian and Behboodi, 2017; Ghasemi et al., 2020). Based
on related works in the respective fields, a systematic simulation-optimization-human
loop is designed, enabling decision makers to identify both weak points and
improvement potentials. An exemplary application of the toolkit for design decisions
in the South-African BSCs is described afterwards. The toolkit uses deterministic
optimization to configure the network structure for the stochastic simulation, which
in turn analyses constraints and changes to the target function in an iterative process
while considering further operational decision-making activities. The aim is to find an
improved supply chain design that performs well under stochastic influences. Thus,
we propose a toolset enabling the depiction and evaluation of the dynamics between
different actors and their decisions throughout a complete BSC. The remaining paper
is structured as follows: The relevant foundations presented within Section 2. In
Section 3, we describe the proposed simulation environment as well as the tools to
consider decision support within the simulation and present an exemplary use case.
The paper closes with a discussion and outlook in Section 4.
2 Foundations
2.1 BSC management
Blood products are not or only partly substitutable resources and people in critical
conditions may depend on their availability (Heidari-Fathian and Pasandideh, 2017).
In order to increase availability, an efficient BSC is required, comprising the
minimization of costs and the avoidance of wastage while meeting the demand. The
BSC process can be divided into four stages or echelons, being: (1) collection, (2)
production (including testing and processing, i.e. separation into different blood
products), (3) storage and inventory, as well as (4) distribution, each containing main
decisions on a strategical, tactical and operational level (Osorio et al., 2015). Some of
those decisions are interdependent, so an isolated approach could lead to a myopic
perspective which makes a real-life application impractical (Osorio et al., 2015).
Verbesserung des Krisenmanagements in Blutlieferketten 179
This work focuses the South African blood transfusion system. There are two blood
establishments operating in South Africa: the Western Cape Blood Service (WCBS),
responsible for the Western Cape province, and South Africa National Blood Service
(SANBS) providing blood transfusion services in the remaining parts of South Africa
(Wise et al., 2020). As nonprofit organizations, WCBS and SANBS must operate cost-
effectively. Both blood establishments collect blood from 100% voluntary, non-
renumerated donors and operate facilities on all relevant echelons of the BSC. SANBS
operates 85 blood banks serving more than 600 hospitals. WCBS runs 7 blood banks
providing blood products to approximately 190 hospitals.
2.2 Simulation-based optimization for blood logistics and
mobile blood donation
While simulation-based systems have commonly been applied to evaluate and design
BSCs, they usually only focus on single echelons and do not enable an integrated view
of the whole BSC (e.g., Alfonso et al. (2012) and Lowalekar and Ravichandran (2010)
focus on the collection stage). Only a few works that consider the whole BSC within
their simulation models exist (e.g., Baesler et al. (2014)). Combining simulation and
optimization approaches is an even rarer methodological approach. For example,
Lang (2009) combine heuristics with a simulation to depict the impact of
transshipments and substitutions in the BSC, while Osorio et al. (2017) describe a
combination of linear programming and discrete-event simulation to improve
production planning in BSCs. However, both approaches focus on specific problems
and are not suited to evaluate a broad selection of practical tasks in BSC operations.
Several publications have addressed the concept of locating mobile facilities or
routing collection vehicles in order to increase the number of blood donations.
Sachdev et al. (2016) for example describe the concept of blood donation mobiles in
India and analyze their use between 2012 and 2014. The authors list several
advantages of using mobiles including the flexibility and the existing air-conditioning
that is especially helpful during hot periods. Şahinyazan et al. (2015) study a mobile
blood collection system in Turkey. Blood mobiles are routed through the area and stay
at least one day at each chosen location, up to three days. Shuttles take the blood
donations to the processing facilities. Gunpinar and Centeno (2016) present an IP
model to simultaneously identify the daily number of blood mobiles to operate while
minimizing the distance travelled. The authors allow the blood mobiles to relocate
during the day and assume a service time of three to seven hours at each location. Two
routing models for blood mobiles are analyzed by Rabbani et al.(2017), Heidari and
Pasandideh (2018) as well as Karadağ et al. (2021) integrate mobile collection units
into an Mixed-Integer Program (MIP) that optimizes the whole BSC. A two-stage
stochastic selective-covering-inventory-routing model was proposed by Bashiri and
Ghasemi (2018) with stochastic blood demand as well as blood donations.
3 Simulation
3.1 Simulation framework
The South African BSC structure is modelled using the HumLog Suite, which was
developed in previous works by the authors (Widera et al., 2017a; Widera et al.,
2017b). HumLog Suite offers an environment for agent-based simulations of
180 Horstkemper, Dennis et al.
humanitarian logistics operations. It is based on the commercial simulation software
AnyLogic and thereby extendable for specific use cases. In addition to AnyLogic,
HumLog Suite offers a predefined set of common humanitarian actors and their basic
decision logic, like beneficiaries representing needs, transport units for goods and
people or medical and storage sites. These actors provide basic parameterization and
state charts that can be specialized for the requirements of the modelled context.
Furthermore, it offers a web-based dashboard to configure simulation scenarios and
to view performance metrics derived through the simulations in a practitioner readable
manner. In the past, HumLog Suite was applied in other contexts like flooding events
and urban evacuation scenarios (Detzer et al., 2016).
In our case, the South African BSC is represented by a set of actors within the agent-
based simulation. Stationery facilities are the collection points, at which blood is
donated by the population, the testing centers, at which the infectious disease testing
is performed, the blood processing center, which transform the collected blood into
the various blood products, the blood banks, which store the blood or blood products
for a period of time, and the hospitals, which administer the blood to patients and
hence create the demand in the system. Each of the facilities is defined by its location
in latitude and longitude, its processing capacities and resource requirements. The
blood banks provide an additional storage capacity and hospitals create the demand
for blood products. Individual facilities can also be situated at the same location,
which eliminates transports between them and affects their cost structure. This is for
example the case, if a hospital runs an own blood bank. The logistics operations
between these facilities are taken care of by a logistics service provider, who offers
services via drones and trucks. These transport units must fulfil specific requirements
for the transportation of blood products, like for example a guaranteed steady cooling.
They are defined by their overall availability at a certain facility, their transport
capacities, costs factors and further parameters on their mode of transportation, like
for example speed. Finally, the potential supply of blood to be donated is modelled in
the population. The supply chain can be distinguished between a supply-driven “push”
and a demand-driven “pull” segment. Starting from collection points via test and
processing centers to the storage in the blank banks, the supply chain is supply-driven
and provides as much blood and blood products as possible according to the
configured donation and processing plan. The amount of blood donations of the
various blood types at each collection point follows a normal distribution with a
configured mean, standard deviation and size for each of the local donor populations.
The hospitals have a demand-driven order policy with the blood banks. The demand
for blood products is derived from a safety stock at the hospitals and the local
population in size and blood type division.
To properly account for the influence of transports in humanitarian operations under
potential crisis events, HumLog Suite utilizes a geographical information systems
(GIS) map based on the real-world road, water and air infrastructure. This
functionality is provided by the AnyLogic software and was customized to be
manipulated according to disaster events, to block parts of the network as consequence
of a flooding, for example. According to the mode of transportation, each transport
unit needs to route within the GIS space, which results in realistic transport conditions.
A specific simulation scenario is defined through a set of input parameters, which
affect the blood product demand, the blood donations, the facility placement, and their
allocation to each other. The location and capacities of hospitals is seen as a long-
Verbesserung des Krisenmanagements in Blutlieferketten 181
term. Yet, disaster events might affect the availability of a hospital due to for example
power outages. Still, within a simulation study, is it typically not changing. The
demand for blood products can be varied to compare best to worst case scenarios and
can be subject to stochastic influences. Also, the blood testing facilities in South
Africa are fixed to two major locations as described before. The amount, location,
capacity, safety stocks and allocation of blood banks to hospitals can the varied to test
different network configurations toward the objectives, e.g., the minimization of
transport costs. The blood processing is defined alike. A set of stationary and mobile
donation sites is provided to collect as much blood as needed to fulfil the demand.
Given these input parameters, the simulation will assess the BSC performance under
potential additional constraints like affected infrastructures or demand fluctuations.
The simulation user interface is displayed in Figure 1.
HumLog Suite tracks each individual transport and the amounts of donated, tested,
produced, stored and administered blood at the locations as well as the resource
utilization over time. This data is used to conclude on key performance indicators for
the BSC, which are defined in line with the objectives. Among others these are the
utilization of blood and blood products, the amount of discarded blood products due
to exceeded shelf life, the transport and overall supply chain costs and the utilization
rates of the facilities and transport units. The analysis outcomes are aggregated and
displayed via the web-interface, which can point practitioners towards current
bottlenecks and improvement potentials to be tested in further simulation scenarios.
Figure 1: Simulation Run
3.2 Simulation and optimization
Following the VDI3633 Part 12 (Simulation and optimisation), two different ways to
connect the simulation of the BSC with optimization-based decision support were
implemented (c.f. Rabe and Clausen, 2015). To perform an optimization of strategic
decisions, the simulation sequentially follows after an optimization of said decisions.
Thus, strategic decisions, e-g. the location of facilities within the BSC, are first
determined by a deterministic optimization, in this case using MIP models and the
commercial CPLEX optimization software. They are then used as an input for the start
of the simulation, which in turn can be used to evaluate the results of the optimization
182 Horstkemper, Dennis et al.
considering stochastic influences such as demand variations and disruptions caused
by external influences such as crises.
In contrast, operational decisions are supported through an optimization integrated
into the simulation. At predefined points in time, the simulation model is paused and
calls the execution of an optimization as a hierarchically subordinate element. This
optimization considers the current state of the simulation, e.g., to perform operational
planning tasks such as route planning or the allocation planning of the blood products
to the different facilities within the BSC. These optimizations can either be performed
using MIP models and the external CPLEX solver again, or by performing manually
implemented heuristics for the respective problems. Depending on the analyzed
timeframes and decision problems, tactical planning problems can be implemented as
either of these optimization types. Figure 2 depicts a sequence diagram visualizing the
two different optimization approaches.
Figure 2: Simulation and Optimization
3.3 Application case: demand-dependent assignment of
mobile blood donation facilities
In addition to stationary blood donation sites, mobile blood donation facilities
(MBDF) are used to make it easier for potential blood donors to reach the donation
sites and thus to increase their willingness to donate. In the South African BSC,
MBDFs operate in a single allocation Hub-and-Spoke system. Therefore, operating
locations for MBDFs are always assigned to only one hub and each MBDF returns to
the hub after operations. Hubs are usually located at blood banks or blood processing
centers, where the collected blood can be processed and/or forwarded. Due to their
decentral nature, usually only whole blood components are collected by MBDF. The
MBDFs visit their operating locations mostly within a fixed pre-planned sequence.
Considering the corona epidemic, two contrasting developments influenced the south
African BSCs. On the one hand, the government’s action to forbid the sales of tobacco
and alcohol severely reduced the need for a number of blood products in the hospitals
(e.g., through the reduction of violent behavior under the influence of alcohol). On the
other hand, the limited availability of public transport, sicknesses and quarantine
measures led to an understaffing of several facilities in the BSC, limiting the blood
Verbesserung des Krisenmanagements in Blutlieferketten 183
supply even further. Thus, the existing sequence for operating the MBDFs is not
aligned to the specific demand for blood products anymore. Furthermore, as not all
MBDFs can be staffed, the efficiency of their utilization gains even more importance.
Hence, one suggestion by domain experts of WCBS was to allocate the MBDFs
considering the current stock and demand situation, taking the distribution of possible
donors and their blood donation characteristics into account as well. Thus, a MBDF
can be placed at a location that offers a concentration of blood donors that can provide
a specific blood product (extracted from anonymized data from past donations), which
is currently out of stock. The simulation-based optimization framework shall hereby
be used to evaluate, this suggestions potential to reduce shortages within the BSC.
Keeping current staff availability in mind, the optimization preceding the simulation
is used to determine the maximum number of available MBDFs, as well determining
the hub locations. Therefore, a capacitated facility allocation problem is used
(Melkote and Daskin, 2001). Depending on the operating radius of a MBDF, the
possible operating locations near the hub location are enumerated and exclusively
allocated to the MBDF. The maximum amount of MBDFs, their hub locations and
their operating locations are then considered as an input for the simulation study itself.
During the simulation, aside from performing the usual transportation, transformation
and storage processes within the BSC, it needs to be decided which operating locations
are visited by the MBDFs on a daily basis. Therefore, the subordinate operational
hierarchic optimization is called at the beginning of each day within the simulation.
As an input for the optimization, the current demands and storage amounts of the
different blood products weighted across all facilities in the BSC are used to determine
which blood products are needed the most. Furthermore, the MBDFs, their possible
operating locations and the expected blood donations per location are considered. The
optimization model is then formulated as an MIP. Within the target function the
demand fulfillment rate is being maximized. Calculating the expected donations using
a specific combination of placements of the MBDFs and matching it to the existing
demands, the fulfillment rates can be determined. To retain cost optimality, the
operating costs of the MBDF are then subtracted from the demand fulfillment rate
using respective weights. This modeling approach was chosen in contrast to the
modeling using penalty costs, as the penalty cost factors itself are difficult to
determine, i.e., cost factors for missing blood products that put patients at risk. The
chosen locations for the operation of the MBDF are then returned to the simulation
and used for the calculation of the blood donations and succeeding processes
throughout the day. This is repeated until the simulations time horizon concludes.
Figure 3: Simulation Results
184 Horstkemper, Dennis et al.
First simulation studies were performed on the South African BSC based on the
recorded data and statistics on blood type characteristics in the South African
population. The objective of these first assessments was to validate the behavior of
the combined simulation and optimization models and to identify first insights on the
current performance of the BSC on highly aggregated values. The left-hand side of
Figure 3 plots the BSC performance over an increasing demand factor across several
simulation runs. With increasing demand, one can assume that in total more MBDFs
are used, which was also observed. It has to be pointed out that the total of MBDFs at
a given point in time was limited to 20 units, which peaked at 18.8 units used in
average. The MBDF Rate depicts the rate of donations in relation to the total amount,
which has a decreasing gradient, meaning that the next MBDF will add less donations
due to the higher costs. While the increasing demand naturally decreases the
shortages, the outages increased until no additional blood donations are produced and
from then on, they decrease as well. This is reasoned by the overproduction of blood
products of common blood types, which cannot be avoided, even when trying to
optimize for donations matching blood type specific demands. With further increasing
demand, shortages are expected to increase again, which was not recorded to this
point. The right-hand side of Figure 3 plots additional performance indicators over an
increasing MBDF count considering a fixed demand rate. Setup costs grow linearly
as expected. Likewise, the linear transport costs are explained by the Hub-and-Spoke
system. As MBDFs only collect whole blood, the whole blood costs of the network
have an increasing gradient when more units are deployed. The MBDF Rate has a
decreasing gradient again, yet not as significant, suggesting further analysis on that
observation is needed in future research. Shortages and Outdates behave quite similar
due to the increased production quantities. Yet, with a fixed demand, the outages are
not expected to decrease as before as overproduction of blood types without demand
will occur. 20 MBDFs seems to be the suggested limit for this demand scenario with
close to zero shortages remaining. Both graphs illustrate a reliable and expectable
behavior of the model supporting the validation of the combined simulation and
optimization. Overall, the current analysis suggests, that a suchlike dynamic allocation
of MBDFs is suitable to reduce shortages in the BSC especially under changing
environmental influences. Limitations are the yet limited extend of simulation studies
and the mixed use of WCBS and SANBS data with general population data, which
can deviate from the individual organization’s reality, necessitating further research.
4 Discussion and conclusion
BSCs are complex networks encompassing many legally independent actors,
transporting a critical good with a very limited availability and shelf life. Nevertheless,
decision support tools known in the general logistics domain are often not being
applied or transferred to this domain. As such, commonly used tools such as
simulation and optimization are not employed often, which is particularly detrimental
in times of crises, were established action plans do not consider the changed
environmental influences onto the BSC. We have proposed a simulation framework
enabling an integration of optimization algorithms for strategic, tactical, and
operational decision support, displaying their interactions and the influence of an
actor’s decision making onto the overall BSC. A suchlike system enables the BSC
operators to adapt and prepare the BSC processes to the dynamically changing
environmental influence, e.g., during a crisis. Furthermore, we gave the example case
Verbesserung des Krisenmanagements in Blutlieferketten 185
of the dynamic operating location choice of MBDFs to illustrate the interactions
between the different optimization types and the simulation. The simulation-based
optimization enables the evaluation of the process to dynamically choose the
operating locations of the MBDFS based on the current stocks of blood products
across the BSC. Nevertheless, the lack of detailed case data does limit the possibilities
to evaluate the overall system at the current point in time. Future work therefore needs
to include studies containing more detailed case data from the individual BSC
operators and facilities to enhance the evaluation of the system. Additionally,
extended workshops and interviews with the stakeholders of the BSC operators need
to be performed to define other relevant evaluation cases, which then can be
implemented and tested also. Finally, performance analysis and improvements need
to be done to ensure the applicability of the approach to support operations in practice.
Acknowledgement
This research was conducted within the project “Blood Information System for Crisis
Intervention and Management” (BISKIT) funded by the German Federal Ministry of
Research and Education in the context of its funding program for International
Disaster and Risk Management (IKARIM).
References
Alfonso, E.; Xie, X.; Augusto, V.; Garraud, O.: Modeling and simulation of blood
collection systems. Health care management science 15 (2012) 1, pp. 63–78.
Baesler, F.; Nemeth, M.; Martínez, C.; Bastías, A.: Analysis of inventory strategies
for blood components in a regional blood center using process simulation.
Transfusion 54 (2014) 2, pp. 323–330.
Bashiri, M.; Ghasemi, E.: A Selective Covering-Inventory-Routing problem to the
location of bloodmobile to supply stochastic demand of blood. International
Journal of Industrial Engineering & Production Research 29 (2018) 2, pp. 147–
158.
Detzer, S.; Gurczik, G.; Widera, A.; Nitschke, A.: Assessment of Logistics and
Traffic Management Tool Suites for Crisis Management. In: European Transport
Conference, 2016,
Ghasemi, P.; Khalili-Damghani, K.; Hafezalkotob, A.; Raissi, S.: Stochastic
optimization model for distribution and evacuation planning (A case study of
Tehran earthquake). Socio-Economic Planning Sciences 71 (2020), pp. 100745.
Gunpinar, S.; Centeno, G.: An integer programming approach to the bloodmobile
routing problem. Transportation Research Part E: Logistics and Transportation
Review 86 (2016), pp. 94–115.
Heidari, H.; Pasandideh, S.H.: Bi-objective optimization of a blood supply chain
network with reliability of blood centers. Journal of Industrial and Systems
Engineering 11 (2018) 3, pp. 190–213.
Heidari-Fathian, H.; Pasandideh, S.H.: Modeling and solving a blood supply chain
network: An approach for collection of blood. International Journal of Supply
and Operations Management 4 (2017) 2, pp. 158–166.
186 Horstkemper, Dennis et al.
Karadağ, İ.; Keskin, M.E.; Yiğit, V.: Re-design of a blood supply chain organization
with mobile units. Soft Computing 25 (2021) 8, pp. 6311–6327.
Lang, J.C.: Production and inventory management with substitutions. Springer
Science & Business Media 2009.
Lowalekar, H.; Ravichandran, N.: Model for blood collections management.
Transfusion 50 (2010) 12pt2, pp. 2778–2784.
Melkote, S.; Daskin, M.S.: Capacitated facility location/network design problems.
European Journal of Operational Research 129 (2001) 3, pp. 481–495.
Osorio, A.F.; Brailsford, S.C.; Smith, H.K.: A structured review of quantitative
models in the blood supply chain: a taxonomic framework for decision-making.
International Journal of Production Research 53 (2015) 24, pp. 7191–7212.
Osorio, A.F.; Brailsford, S.C.; Smith, H.K.; Forero-Matiz, S.P.; Camacho-
Rodríguez, B.A.: Simulation-optimization model for production planning in the
blood supply chain. Health care management science 20 (2017) 4, pp. 548–564.
Rabe, M.; Clausen, U.: Verknüpfung von Simulation und Optimierung: Kategorien
und Beispiele‐Ein Bericht über die VDI Richtlinie 3633 Blatt 12 (2015).
Rabbani, M., Aghabegloo, M., Farrokhi-A,H.: Solving a bi-objective mathematical
programming model for bloodmobiles location routing problem. International
Journal of Industrial Engineering Computations 8.1 (2017), pp 19-32.
Ramezanian, R.; Behboodi, Z.: Blood supply chain network design under
uncertainties in supply and demand considering social aspects. Transportation
Research Part E: Logistics and Transportation Review 104 (2017), pp. 69–82.
Sachdev, S.; Singh, L.; Marwaha, N.; Sharma, R.R.; Lamba, D.S.; Sachdeva, P.:
First report of the impact on voluntary blood donation by the blood mobile from
India. Asian journal of transfusion science 10 (2016) 1, pp. 59.
Şahinyazan, F.G.; Kara, B.Y.; Taner, M.R.: Selective vehicle routing for a mobile
blood donation system. European Journal of Operational Research 245 (2015) 1,
pp. 22–34.
Stanworth, S.J.; New, H.V.; Apelseth, T.O.; Brunskill, S.; Cardigan, R.; Doree, C.;
Germain, M.; Goldman, M.; Massey, E.; Prati, D.; others: Effects of the COVID-
19 pandemic on supply and use of blood for transfusion. The Lancet
Haematology (2020).
Widera, A.; Konradt, C.; Böhle, C.; Hellingrath, B.: A multi-method simulation
environment for humanitarian supply chains. In: 2017 4th International
Conference on Information and Communication Technologies for Disaster
Management (ICT-DM), 2017a, pp. 1–8.
Widera, A.; Lechtenberg, S.; Gurczik, G.; Bähr, S.; Hellingrath, B.: Integrated
Logistics and Transport Planning in Disaster Relief Operations. In: 14th
International Conference on Information Systems for Crisis Response and
Management, 2017b
Wise, R. D., Gibbs, M. W., & Louw, V. J.: Lockdown and our national supply of
blood products. SAMJ: South African Medical Journal, 110(5), (2020).
