Content uploaded by Ning Wang
Author content
All content in this area was uploaded by Ning Wang on Mar 27, 2021
Content may be subject to copyright.
20 IEEE TRANSACTIONS ON TECHNOLOGY AND SOCIETY, VOL. 2, NO. 1, MARCH 2021
“As It Is Africa, It Is Ok”? Ethical Considerations of
Development Use of Drones for Delivery in Malawi
Ning Wang
Abstract—Since 2016, drones have been deployed in various
development projects in sub-Saharan Africa, where trials, tests,
and studies have been rolled out in countries, including Tanzania,
Uganda, Rwanda, Malawi, Ghana, and the Democratic Republic
of the Congo. The use cases of drones vary, ranging from imagery
collection to transportation of vaccines, lab samples, blood prod-
ucts, and other medical supplies. A wide range of stakeholders
is involved, including governments, international organizations,
educational institutions, as well as industry. Based on a field study
conducted in 2019, this article investigates how drones are used
for medical supply delivery in Malawi—a country where the com-
munity is underserved for healthcare and related infrastructure
underdeveloped, while airspace is largely open and regulations
generally relaxed. The objective of presenting this case study is
to contribute to the evidence regarding the rapid deployment of
medical cargo drones across the African continent, and to spark
critical reflections over the utility, suitability, and impacts of
incorporating drones in the existing health supply chain systems
in resource-poor settings. The discussion revolves around two
aspects: 1) the emergent “African Drone Rise”—is it ok “as
it is Africa”? and 2) the normative role of technology in the
aid sector—is it “a solution looking for a problem”? In con-
clusion, a call for more structured guidance for the systematic
examination and evaluation of the medical cargo drone case is
raised.
Index Terms—Health supply chain system, humanitarian
drone, medical cargo drone, medical supply delivery, public
interest technology.
I. INTRODUCTION
ACCORDING to the World Bank, one-third of the
world’s population lacks regular access to essential
medicines [1]. The volume and complexity of global aid
programs has spawned extensive and complex health sup-
ply chains, many navigating difficult conditions in low- and
middle-income countries with poor infrastructure, complex
logistics, and lack of resources [2]. Although the humanitarian
supply chain represents 60%–80% of humanitarian expendi-
tures, last mile delivery (LMD) logistics is a critical constraint
preventing medical supplies from reaching remote areas, caus-
ing reported vaccines supplied to parts of the targeted countries
to expire before they can be administered [3]. To address the
Manuscript received January 18, 2021; revised February 8, 2021; accepted
February 8, 2021. Date of publication February 11, 2021; date of current ver-
sion March 11, 2021. This work was supported in part by the Swiss Network
for International Studies (SNIS) under Grant C18006, and in part by the Swiss
National Science Foundation (SNSF) under Grant P1ZHP1-181401.
Ning Wang is with the Institute of Biomedical Ethics and History
of Medicine, University of Zurich, 8006 Zurich, Switzerland (e-mail:
ning.wang@ ibme.uzh.ch).
Digital Object Identifier 10.1109/TTS.2021.3058669
LMD challenge, the use of drones is considered useful to help
optimize the health supply chain, due to their technical ver-
satility, operational viability, and economic accessibility [4].
Since 2014, several projects have demonstrated a proof of con-
cept of using drones for medical delivery [5]–[7]. Drone-based
healthcare projects have emerged rapidly across the globe
with a broad range of applications, especially prominent in
sub-Saharan Africa.
Although cargo drones used for development and health-
care purposes are still an emergent approach, there is a wide
array of actors involved ranging from tech start-ups to logis-
tics companies, many partnering with universities, NGOs, and
international aid organizations [3]. Yet due to the early stage
of implementation, to date, little real-world experience or
primary data related to technology performance, operations,
health impact, cost, or acceptability are available [5]. In addi-
tion, the trend of the “African Drone Rise” in recent years,
whereby drones and Africa are being construed as solutions to
each other’s problems, opens up critical questions with respect
to the ethical and societal implications of using drones in the
aid sector [9], [10]. This article constitutes one element of
a research project that examines technological innovation in
the aid sector and how it intersects with moral values, norms,
and commitments [11], [12]. As part of a set of field studies
of different uses of drones by international aid organizations,
this case study investigates the use of drones for medical sup-
ply delivery in the lake area of Malawi. A detailed narrative
account of the case study was presented in [4].
The current paper draws upon the empirical findings of the
field study to develop an analysis with the goal of identify-
ing contextualized ethical considerations, and illuminating the
wider debate about how ethical technological innovation in
the aid sector can be operationalized. This article comprises
three parts: 1) a short summary of a case study of a medi-
cal drone delivery project, including a detailed presentation
of research methods; 2) an in-depth analysis of six cate-
gories of challenges that emerged in the context of the case
study: a) human and environmental safety; b) cargo safety;
c) operational costs; d) infrastructure gaps; e) local capac-
ity; and f) donor dependence; and 3) a discussion around the
phenomenon of the African Drone Rise, and the observed men-
tality of “solutionism,” in light of the concerns and critiques
raised by scholars and practitioners in the field. In conclu-
sion, I call for a prudent attitude in adopting novel technology
in the aid sector and argue that proposals for actionable eth-
ical standards to guide sector-wide innovation practices are
needed.
2637-6415 c
2021 IEEE. Personal use is permitted, but republication/redistribution requires IEEE permission.
See https://www.ieee.org/publications/rights/index.html for more information.
Authorized licensed use limited to: MAIN LIBRARY UNIVERSITY OF ZURICH. Downloaded on March 27,2021 at 02:49:49 UTC from IEEE Xplore. Restrictions apply.
WANG: “AS IT IS AFRICA, IT IS OK”? ETHICAL CONSIDERATIONS OF DEVELOPMENT USE OF DRONES FOR DELIVERY IN MALAWI 21
II. CASE STUDY
A. Drone Delivery Project
In 2017, the Government of Malawi (GoM) established the
Humanitarian Drone Testing Corridor, in collaboration with
a specialized UN agency, where studies could be implemented
on using drones to facilitate LMD and integrating drones
in an optimized health supply chain system in low-resource
settings [5]. Since then, Malawi has become a popular site
for drone testing and pilot projects [4]. In this case study,
a donor agency attempted to find a solution to assist the GoM
with the LMD challenge by transporting lab samples from
the two islands in Lake Malawi—Likoma and Chizumulu—
to the mainland, where land-based health logistics established
by the Ministry of Health (MoH) are routinely used [4]. An
organization, which is referred to as Y in our study, works in
the area of public health and international aid, and is involved
with selecting, procuring, and delivering medicines for infec-
tious diseases, such as HIV/AIDS, in Africa, Asia, and Latin
America [4]. As drones showed potential to provide medical
supplies to hard-to-reach areas, Y was eager to test if they
could include drones in health logistics, and secured funding
for a feasibility study [4]. The study was contracted to a com-
pany with the objective to use drones to connect the health
centers on the two islands to the mainland as a complementary
supply chain solution [4]. Prior to the utilization of drones, the
lab samples were transported by a passenger boat which oper-
ates twice a week and takes about 6–8 h each way; with the
drone, multiple deliveries per day can be performed, taking
about just an hour each way, helping reduce the transportation
time to a large extent [4]. In July 2019, upon receiving the
final approval from the Department of Civil Aviation (DCA),
the project entered into full operation [4]. According to the
drone pilot interviewed, between July and October 2019, over
200 flights were operated for more than 90 h and delivered
about 45 kg of medical commodities over about 30000-km
flight route [4]. During this period, the project undertook rou-
tine deliveries, reducing sample turnaround time from 5 to
8 weeks with the boat to less than 4 weeks with the drone [4].
The project was perceived as technically successful as it show-
cased the state of the art of the technology, and achieved
its intended outcome with respect to lab sample transporta-
tion during the operation. However, there were also ethical
challenges encountered in terms of safety, health impacts, cost-
effectiveness, capacity, and sustainability [4]. An analysis of
these challenges is provided in Section III.
B. Research Methods
1) Research Design: The case study was carried out in
Malawi in October 2019 and February 2020, as part of a larger
program investigating “value sensitive humanitarian innova-
tion (VSHI)” and consisting of multiple case studies. The
qualitative, interview-based research was conducted within
a constructivist paradigm, in which human experience is under-
stood as subjective, local, socially and experientially based,
and culturally and historically specific [13]. Study design drew
upon case study methodology, and we employed qualitative
description as our methodological framework, which aims to
gain first-hand knowledge of stakeholders’ experiences, and
describe their views and perceptions of a particular topic in
a language similar to their own [14], [15]. Our rationale of
using a qualitative description in this case study is to stay
especially close to the data itself, developing a low-inference
analysis by directly organizing and synthesizing data without
further interpretation [15].
2) Participant Recruitment: We acquired and received the
Ethics Review Board approval from the National Commission
for Science and Technology of Malawi on 13 November 2019.
We then recruited interview participants using two approaches.
1) Guided by an interview plan, we sent email invita-
tions to targeted stakeholder representatives, which were
jointly identified by the research team and our local
partners. We recruited 12 individual participants, includ-
ing three development workers (at international and
national levels), one drone technician, six government
officials (at national ministerial and community levels),
and two health workers (at community level). We also
convened a focus group of local community members
(22 villagers in total). All of these participants were
involved in, experienced, or witnessed the drone delivery
project.
2) Following a snowball sampling logic, we recruited
eight further participants by recommendations of
previous participants, to expand the scope of the inves-
tigation and gain complementary and contextualized
data.
The further recruitment included one elected official,
five additional health workers, and two local drone techni-
cians. All of these participants were situated at the community
level and were directly involved in the drone project. The final
sample of 42 participants thus consisted of a diverse represen-
tation of stakeholders engaged in the drone delivery project,
consistent with our goal of maximum variation sampling in
order to explore the common and unique perspectives on the
subject [16].
3) Data Collection: Our main data collection was semi-
structured qualitative interviews. All participants provided
written informed consent prior to their interview. Other
data sources included texts (e.g., publicly available reports),
and observations by the author (e.g., how interviewees inter-
acted with each other), who carried out the field study in
Malawi as the principal researcher. Two types of interviews
were conducted: 1) those that followed an interview guide,
which was developed prior to the field study with targeted
questions centered around key areas of ethical concerns, was
initiated based on expert knowledge and 2) open discussions
to explore topics which emerged from the earlier interviews
(e.g., the regulatory environment in Malawi, or the particulari-
ties of medical drone delivery in the African context). Overall,
12 interviews were conducted between 24 October 2019 and
12 February 2020, including eight individual interviews with
a single participant, three panel interviews with 2–5 partici-
pants each, and one focus group interview with 22 participants.
All interviews were conducted face-to-face and were audio-
recorded, ranging from 46 to 169 min in duration (average
87 min). The individual and panel interviews were conducted
Authorized licensed use limited to: MAIN LIBRARY UNIVERSITY OF ZURICH. Downloaded on March 27,2021 at 02:49:49 UTC from IEEE Xplore. Restrictions apply.
22 IEEE TRANSACTIONS ON TECHNOLOGY AND SOCIETY, VOL. 2, NO. 1, MARCH 2021
TAB LE I
CATEGORIES OF ETHICAL CHALLENGE
in English; and the focus group interview with the community
members was conducted in the local dialect, with translation
provided by our local partner.
4) Data Analysis: Data analysis consisted o two steps:
1) generating validated descriptive summaries of the interviews
and 2) extrapolating thematic categories for further analysis.
For the first step, we developed a written synopsis of each
interview, based on the interview recordings and with refer-
ence to the available texts and observations, with the support
of a research assistant. These synopses were compared with
interview notes taken by the author during each interview,
leading to a comprehensive descriptive summary of the inter-
views. The summary was then sent to anonymous reviewers,
who were involved in the drone delivery project, to ensure that
it was factually accurate. This resulted in a detailed narrative
account of the case study, which was presented in [4]. For the
second step, core concepts were first developed by themes,
and these themes were then clustered in six categories and
extrapolated for analysis. Based on this inductive analysis, four
aspects of ethical considerations were drawn in light of two
key concerns: 1) the emergent African Drone Rise—is it ok “as
it is Africa”? and 2) the normative role of technology in the aid
sector—is it “a solution looking for a problem”? These find-
ings then led to the conclusion of lessons learned from three
aspects, namely, the technical, impact, and benefit-sharing per-
spectives, summarized in Table I. It is worth noting that these
themes represent a rather broad spectrum of challenges, includ-
ing technical aspects such as safety, and operational aspects
such as costs. Some of these aspects may not be ethical issues
per se such as infrastructure or capacity, or may not exclu-
sively relate to drones, such as patient data, but may nonethe-
less cause unintended consequences of ethical concerns, and
are in that sense considered as having an ethical impact. These
categories do not suggest a specific order of criticality and are
presented with a horizontal logic for ease of reading.
III. ANALYSIS OF ETHICAL CHALLENGES
While no severe ethical tensions, tradeoffs, or dilemmas are
observed in this case study, there are ethical considerations
Authorized licensed use limited to: MAIN LIBRARY UNIVERSITY OF ZURICH. Downloaded on March 27,2021 at 02:49:49 UTC from IEEE Xplore. Restrictions apply.
WANG: “AS IT IS AFRICA, IT IS OK”? ETHICAL CONSIDERATIONS OF DEVELOPMENT USE OF DRONES FOR DELIVERY IN MALAWI 23
related to safety, operationality, and sustainability among the
involved stakeholders. In the following sections, I provide
a detailed analysis of how these themes were manifest in the
case study, with references to the wider academic literature.
A. Safety
1) Human and Environmental Safety: First, the biggest risk
for drone operation is the drone technology itself. The drone
technology has inherent technical limitations—drones are bat-
tery powered, and the more powerful the battery, the longer
the distance, but also the heavier the payload, and the big-
ger the drone, hence the more severe safety consequences in
the case of a drone crash [11]. Take the drones used in this
project for example: they can weigh up to 16 kg, can carry
cargo up to 6 kg, and can take off and land in relatively small
spaces without requiring infrastructure beyond visual line of
sight (BVLOS) [17]. For short distances, or where no pickup
is required at the destination site, the drone has the capability
to drop packages without landing by hovering and releasing its
cargo in a box. For longer distances, or where sample pickup
is needed, the drone will need to land for a battery charge. In
the case where the drone crashes, there will be a serious risk
to human safety when flying over populated areas, or causing
environmental damage when flying over the lake. According to
the interview participants, at the time of the operation, Malawi
did not have official safety regulations, which raised concerns
for some stakeholders at the MoH. Although the operations
team took precarious safety measures, it was reported that
there had been accidents both within and outside of the Drone
Corridor, and due to either mechanical or human errors [4].
As the actual drone operations were not executed by the GoM,
but by the drone manufacturers or service providers asso-
ciated with international aid agencies, the safety challenges
lie with them to overcome. Compared to manned aircraft,
unmanned aircraft regulation is either nonexistent or fairly
relaxed currently, making safety issues, including physical and
environmental aspects worrisome.
Another factor that causes safety concerns is the connec-
tivity solution. According to experts, uninterrupted commu-
nication between the drone and the control station is of
paramount importance to ensure safety, as the drone needs
to be monitored constantly and precisely from the ground
station [18]. If the local mobile network is sufficiently estab-
lished, then the drone can be continuously controlled using
mobile data that covers the entire flight route. If such an infras-
tructure is unavailable, then a customized communication plan
needs to be set up, such as a combination of mobile network
and satellite. In this project, the drones used in the opera-
tion are semi-automatic and can vertically take-off and land,
i.e., humans remotely control take-off and landing via mobile
network, and then the control is surrendered to the onboard
computer during the flight. When connectivity is lost, the
drones would remain at its last known position and circle
until connectivity would recover, during which time there is an
increased likelihood of a crash or a so-called “hard-landing,”
e.g., affected by wind or other weather conditions. Since the
poor communication infrastructure in Malawi caused network
interferences to drone navigation, the operations team had to
triangulate between the GSM/LTE network with automatic
switch to satellite communications, when the former is absent,
to make sure that they would not lose the drone mid-air. As
the study indicates, this raised concerns among drone pilots
and involved community members that, if a drone would lose
connectivity when flying over the lake and fall into the water,
it would not only be difficult to locate and retrieve it, but car-
gos, including samples or medicines, would get lost. Although
satellite communications are the most reliable and with rela-
tively low costs in initiating the drone operations, technically
they are not the most optimal, and economically they lead to
higher costs during the operation in the long run [18]. This
challenges the proposed scalability of drone delivery in areas
where communication infrastructure is already fragile.
A third factor is the weather conditions. There are gen-
erally strong and lasting headwinds over Lake Malawi [4].
Prior to the operation, the MoH had warned the project team
about the unfavorable weather conditions, and urged them to
take a cautious approach to verify the capability of the drone
to withstand strong winds, as well as to assess any poten-
tial environmental risks related to it. During the operation,
technical staff at the ground station would check the weather
and assess the air speed of the drone, on the basis of which
they would calculate the wind speed and decide whether they
should abort the flight mid-air. Still, the wind led to one crash,
causing the drone to fall directly into the lake. Weather con-
ditions during a flight can impact the success of a drone
mission, as inclement weather conditions or significant dif-
ferences in ambient temperatures may cause drones to lose
their functionality [19], [20]. The current state of technology
development does not allow the drone to calculate the flight
time itself, whereby it decides whether the fight time is more
than the maximum allowed flight time and to automatically
return. Granted that tracking systems may help locate the area
where the drone fell, and with the lay knowledge about the
current and waves of the lake, it may be possible to roughly
determine where the drone has gone. Granted too that there
may be ways of minimizing the potential risk of environmental
damage caused by cargo leakage, through secured cargo pack-
aging protocols. These risk mitigation measures, nonetheless,
do not reduce the inherent risk of drone accidents caused by
adverse weather conditions, or interrupted communications as
aforementioned. This raises the question as to whether drones,
especially smaller ones, suit the windy environment around
Lake Malawi.
A final factor is safety insurance which relates to all the
above three factors. It is important to note the fundamental
difference between manned or commercial aviation flights and
unmanned or drone flights with regard to insurance. In the for-
mer case, passengers actively accept the risk when they board
the flight, while in the latter case, those on the ground do
not have the opportunity to accept the risk prior to a drone
operation [21]. While the capability of drones in carrying pay-
load or handling wind is critical, it is equally important to
establish insurance policies to protect populations in the oper-
ation areas. According to the participating interviewees at the
MoH, ever since the first drone operator that came to Malawi
whose drone “crashed within the shortest possible time and
Authorized licensed use limited to: MAIN LIBRARY UNIVERSITY OF ZURICH. Downloaded on March 27,2021 at 02:49:49 UTC from IEEE Xplore. Restrictions apply.
24 IEEE TRANSACTIONS ON TECHNOLOGY AND SOCIETY, VOL. 2, NO. 1, MARCH 2021
even injured a villager,” they have witnessed constant changing
of drones from one model to another, none with an insur-
ance policy. They are of the view that as the Malawians never
demanded drones, they are largely unaware of potential safety
risks related to drones and need protections from the govern-
ment. According to our interviews, expectations from both the
health and the aviation sectors to establish drone insurance,
ranging from injury of individuals to damage of leaked medi-
cal supplies, is currently high in Malawi. In the absence of an
insurance policy in Malawi, the GoM and the respective drone
operators opted for signing a mutual agreement to settle the
safety insurance aspect of the drone projects in question. Some
stakeholders seemed to be inclined to downplay the ethical
risks with a reductionist view, focusing solely on the “ethical
conduct of business,” e.g., no bribery to regulatory author-
ities, or no dealings with weaponized drone manufacturers.
However, the often-overlooked unintended harmful conse-
quences associated with the introduction of new technology
may, in fact, be much more detrimental to vulnerable popu-
lations and their living environment, no matter how “ethical”
the conduct of the business may be.
2) Cargo Safety: With respect to lab sample packaging,
Malawi takes an approach whereby every blood sample taken
is considered infectious until it is tested negative. To ensure
cargo safety, the MoH adopts a so-called “three-layer” system.
The first layer varies depending on the samples—in the case of
a TB sample, the sputum is packaged in bottles with a screw
top; in the case of an HIV sample, as it is dried blood sam-
ples on absorbing papers, it is packaged in envelopes with
seals. These samples are then kept in Ziploc bags, which is
the second layer. They are then put in a bigger container made
of hard plastic locked with a screw top, which is the third
layer. This packaging system is designed to ensure that the
person carrying the samples is protected, as is the environ-
ment and the bystanders. In this project, the fully packaged
samples are placed inside the cargo box, which is made of
fiberglass and is attached to the bottom of the drone during
the flight [17]. In principle, even if a drone would crash, the
cargo box would remain intact, and the risk of sample spillage
or leakage should be minimal. However, according to both the
technical and health personnel interviewed, as the cargo box
itself is not locked and can be opened with a magnet, in the
case of a drone crash, the sealed bottles or envelopes may still
fall out of the cargo box and be dispersed (although the actual
samples may not be revealed unless someone picks them up
and opens them). Since the drone pilots do not typically handle
the samples, other than keeping track of what is being carried
by the drone, and logging how many samples and which types
of samples are being loaded into the drone, the drone opera-
tors essentially provide a “postal service” using drones instead
of ground vehicles. This leaves questions of responsibility for
safety risks more complex, if and when there is sample leakage
in the case of a drone accident.
The importance of cargo safety related to dangerous goods
in humanitarian aid and emergency response is strictly regu-
lated by the International Civil Aviation Organization (ICAO)
in its 2020 Guidance. In this document, ICAO focuses specif-
ically on the safety risks, responsibilities, and mitigation
measures related to potentially dangerous cargos carried by
drones [22]. The Guidance also provides technical definitions
of “dangerous goods,” and recommendations regarding their
handling, including detailed guidelines laid out in its four
Appendixes. According to the Guidance, examples of danger-
ous goods include infectious substances such as lab samples
for analysis, and toxic substances such as certain medicines or
chemical, among others [22]. It also affirms that in the case
of biological substances, pathogen data sheets or information
about the hazards of infectious substances, including deactiva-
tion and waste disposal, should be made available [22]. Given
that some of the cargo transported in this project was infectious
samples, such as TB sputum, as well as other potentially dan-
gerous chemical supplies, such as lab reagents used to run the
sample testing, losing cargo became the second biggest safety
concern of the health personnel across the MoH. The ICAO
Guidance classifies risks associated with the dangerous goods
transported by drones by levels, and recommends risks to be
assessed in relation to the consequence of their effects if they
are released [22]. It asserts that, in addition to normal flight
risks associated with operating routes, obstacles, altitudes, or
take-off and landing areas, special efforts must be made when
these goods are transported over populated areas, remote areas,
or environmentally sensitive land and waters [22]. As govern-
ments around the world are currently working on creating and
adapting legislation to ensure both safety and development
related to the humanitarian use of drones, the regulatory land-
scape evolves rapidly [3]. Although international development
work may not always take place in emergency situations, the
safety standards should still apply.
Closely related to sample packaging is the issue of patient
data. According to the interviewed health personnel involved
in sample packaging, the patient data related to the lab samples
in this project is personally identifiable, as the samples have
both patient name and a unique ID number on them. In addi-
tion, this data is also demographically identifiable, as there is
a facility code on the sample, as well as a laboratory requisi-
tion form which describes the test that needs to be performed
on the sample. This means that if the sample is revealed, it is
possible to link it with a particular patient who is associated
with a particular health facility. Patient confidentiality, there-
fore, is at risk of being compromised in the case of drone
accidents. In the U.S., for instance, medical drone delivery
operations must comply with the Health Insurance Portability
and Accountability Act (HIPAA), which is a set of regulations
established to protect the confidential and private information
of patients [20]. According to the Act, HIPAA may be vio-
lated if individuals who are not involved in the direct care
of patients view patient information on the labels of medical
specimens or medications [20]. In this project, the cargo to
be transported by the drone is entrusted to the involved health
and technical personnel. As the drone pilots receive the sample
in packages ready to be loaded into the drone, patient data is
handled separately from the drone operation and is collected
and processed within the involved health facilities. Concerns,
however, arose around potential data leakage as a result of
a drone accident. There may be a perception that patient con-
fidentiality risks are not drone specific, as they also apply to
Authorized licensed use limited to: MAIN LIBRARY UNIVERSITY OF ZURICH. Downloaded on March 27,2021 at 02:49:49 UTC from IEEE Xplore. Restrictions apply.
WANG: “AS IT IS AFRICA, IT IS OK”? ETHICAL CONSIDERATIONS OF DEVELOPMENT USE OF DRONES FOR DELIVERY IN MALAWI 25
traditional health supply chain solutions. Yet, it is noteworthy
that there is a heightened risk of unintended harmful conse-
quences caused by the frequency and height, as well as the
scope of reach, of drone flights, compared to ground vehicles
as the latter depend less on parameters, such as weather and
mobile communication. This suggests that downplaying the
severity of patient data risks associated with cargo drones is,
at the least, an imprudent attitude.
B. Operationality
1) Infrastructure Gaps: Despite the media hype and indus-
try excitement, drones are not the silver bullet to the health
logistics system. For example, in emergency cases where
patients on the islands need to be attended, unlike ambu-
lances, drones cannot assist in the transportation of humans.
Another example is the case of a drone crash, in which sit-
uation the drone-based health supply chain would crash as
well. According to the DCA, after a drone crash, the operation
must be halted immediately while the technical team return-
ing to the Drone Corridor, undertaking thorough technical
checks, demonstrating revised flight strategy, and reapplying
for approvals—a lengthy process different from traditional
logistic solutions, such as boats or motorbikes. As happened in
this project, routine sample delivery got interrupted when the
drone crashed during the operation, and the local health facili-
ties had to rearrange logistics, i.e., going back to the passenger
boat. These aspects imply that to maintain a stable delivery
service, the health facilities on the islands cannot rely solely
on the drone, but would need an alternative that is somewhat
more reliant, and that they can safely fall back on. As the study
indicates, drones were useful to help with sample transporta-
tion with an improved health outcome; the problem occurs
when the operation got interrupted by, e.g., weather, connec-
tivity, or drone accidents, in which cases the improvement
achieved would be setback. The real issue here is not about
choosing between drones or ground vehicles, but to ensure
reliable and continued health logistic services. What remains,
in the long run, is the underlying structural problems of the
impoverished public health system in Malawi—had the GoM
had more systemic solutions to resolve the LMD challenge in
the health sector, drones would have unlikely appeared in their
agenda in the first place. The dilemma is that such alternatives
are not immediately present to Malawi.
Another challenge relates to lab sample processing,
which concerns the potential overburdening—as opposed to
strengthening—of local health systems. The main factor for
sample processing is the turnaround time, which is measured
by the total running time (TRT) and classified as T1 (sample
collecting), T2 (sample processing), and T3 (test result dis-
semination). In theory, even though the use of drones does
not change T2, the TRT still ends up shorter as a result of
the reduced time of T1 and T3. Take the Central Molecular
Lab (CML) of Malawi for example: in this project, although
the CML received the same number of samples in total, the
frequency of delivery has changed with the drone. This, to
some degree, helps the lab to reduce the TRT, as there is now
a more stable sample inflow compared to the past. On the flip
side, however, there are other variables at play, which actually
led to a prolonged TRT on some occasions. This is because,
since the drone delivery started, both the lab personnel and
equipment turned out to be insufficient, although the CML had
an increase of staff and a second platform installed in the lab.
With the two platforms, the lab personnel had to work days and
nights on double shifts to keep up with the increased inflow
of samples. Additionally, although the TRT has gone down
with a constant sample supply, reality suggests that with the
increase of samples, the lab has been significantly hampered
by insufficient consumables, e.g., lab reagents for biochemistry
tests or full blood counts. This is because Malawi has the so-
called “push system” in health consumables supply, and the
ongoing supply has not been adapted to the sudden demand
for lab consumables and is underdelivered. Furthermore, the
increased inflow of samples also hampered the machines them-
selves due to overworking, which caused some downtime to
the lab. As a result, to cope with the drone deliveries, the lab
would need more equipment to account for the number of sam-
ples coming in, and the same has to apply to the number of
staff and consumables. Looking from the big-picture perspec-
tive, drones offer a partial solution to a much more complex
problem.
The sample processing aspect is closely related to the health
facility capacity in remote areas of Malawi. Take HIV patients
for example: many who need to be treated for HIV are also
infected with TB. If a health facility needs to do an HIV viral
load test, they also should do a TB culture to understand how
the drugs are working to help with the management of the
same HIV patient. Unfortunately, donors such as those in this
project, have a desired focus on HIV, overlooking other interre-
lated causes that contribute to the patient’s overall conditions.
Presently, each district hospital in Malawi has a chemistry lab
that can do geneXpert, but only 3–4 labs across the whole
country can handle TB culture for a drug resistance case.
Yet, it is not immediately feasible for the MoH to drastically
improve lab facilities in remote areas, which would allow for
local testing of samples. Drones offer the MoH a solution to
the LMD challenge they are faced with, and give a sense of
relief to the health personnel on the islands. To overcome the
infrastructure barriers, the GoM implemented short- and long-
term measures. While long-term measures inevitably suggest
more budget at the central level, short-term measures give the
districts flexibility to use locums to cover the staff shortage.
Nevertheless, as Malawi already suffers from a general short-
age of health personnel, this measure hardly helps. As such,
just like the case of the CML, even if the MoH would be able
to install additional lab equipment in remote health facilities,
the lack of human resources is still a barrier hard to over-
come, let alone the lack of electricity and other material such
as the aforementioned lab consumables in those areas. The
question, hence, boils down to how to compensate the poor
infrastructure with an already-burdened health system.
2) Operational Costs: From the investment perspective,
the drone operation is an expensive venture as, to date,
most drone manufacturers are in start-up modes and are
still in the stages of research and development in the tech-
nology. The capital cost needed for operation is high, and
the maintenance costs, including infrastructure and manpower
Authorized licensed use limited to: MAIN LIBRARY UNIVERSITY OF ZURICH. Downloaded on March 27,2021 at 02:49:49 UTC from IEEE Xplore. Restrictions apply.
26 IEEE TRANSACTIONS ON TECHNOLOGY AND SOCIETY, VOL. 2, NO. 1, MARCH 2021
are even higher, requiring a massive amount of resources
to be invested [5], [8], [17]. Traditionally, supply chain man-
agement systems account for the costs of device operation
and maintenance, as well as transport time, road condition,
warehousing, and staff [23]. The cost of adding drones to
the supply chain will be determined not only by these mea-
sures but also the unique drone-specific considerations [8].
According to an FHI360 study, understanding the conditions
under which drones are cost effective is critical, but it is a com-
plex area for investigation [23]. First, the costs include both
the cost of the technology and the cost of service provision.
These cost models vary substantially depending on the specific
system used as well as staff training to operate and maintain
them [23]. Second, weather conditions are a big factor in drone
operating costs. Knowing the impact of wind, humidity, eleva-
tion, precipitation, and temperature on supply chain operations
will be critical [23]. Third, in addition to direct costs gener-
ated by the drone service, opportunity costs also play a role.
On a system level, it is difficult to measure the value of faster
turnaround times for lab test results, against the value of hav-
ing a health provider remain in a facility for a day rather than
transporting medical goods [23]. Presently, the cost effective-
ness of using drones in the development context is largely
unknown [5], [8]. Many donors across the globe have, there-
fore, kept a vigilant attitude toward investing in medical cargo
drones.
From the beneficiary perspective, for resource-constrained
countries like Malawi, the only way to run drone operations is
relying on donor funds. In the case of Malawi, the drone opera-
tions do not generate an additional budget for the health sector,
and the MoH on its own cannot afford drones. In general, there
is the question of how money should be spent in the health sec-
tor. For most health personnel interviewed in our study, there
is an overwhelming preference for improving the health facil-
ity capacity in remote areas, to “flying this small thing around
Malawi,” as some health workers sarcastically noted. They
are of the view that rather than using the precious resources
allocated to public health on drone operations, Malawi would
be better-off procuring new lab equipment, or training more
health workers in the long run. Not only do they perceive
the drone operation as wasting resources but also the result
of improper agenda setting at the central level. For the MoH,
however, although drone operations cost a larger sum of funds,
if it indeed proves to be cost effective in bringing better health
outcomes, they are willing to seek donor funds and justify
costs. In reality, due to the above-mentioned lack of knowl-
edge on the actual costs of drone operations, most donors are
currently still interested in testing a proof of concept by invest-
ing in a pilot study to explore the operational feasibility. These
projects tend to be short term from a few days to a few weeks,
of experimental nature, with predefined objective and method-
ology, and managed by international operations teams [4].
Although the future of medical cargo drones is promising, the
immediate challenge comes down to the cost of running the
service on the ground in resource-constrained environments.
Hence, the beneficiary governments need to carefully review
the long-term costs to assess if drones indeed make sense to
their particular contexts.
The challenge on the system level also relates to the so-
called business model of medical cargo drones across Africa.
As the current practices suggest, there are different donors and
healthcare providers independently conducting drone opera-
tions in different countries [5], [8]. From a knowledge build-
ing perspective, some of these projects can be complementary
to each other, in terms of types of drones used (parachuting
or vertically take-off/landing), modes of delivery chosen (one-
way dropping or bidirectional pick-up/dropping), flight routes
defined (long- or short-distance), etc. [5]. Still, working in silos
may potentially create competition over attention, resource,
agenda setting, and even the airspace. Countries like Malawi,
where health systems are already fragile, need a healthy and
sustainable environment to incubate innovation. In addition,
drone development, impact evaluation, and final implementa-
tion require longer-term investment with funds that go beyond
pilot projects [5], [17]. This leaves a large investment space
for the private sector, where venture capital funds are avail-
able to enable the technical robustness needed. However, it is
worth noting that cost-effectiveness analysis may not be the
top priority for such investors, as they may not be seeking
cost-effective interventions; but rather, are keener on finding
out how to operate drones on the ground in the development
space, and how to integrate this technology into an existing
supply chain while the operational costs continuing to decrease
as the technology further matures [24]. Ultimately, it is up to
the national governments to assess their needs, conduct cost-
effectiveness analysis, and gauge in between various tradeoffs
that are particular to their own contexts.
C. Sustainability
1) Local Capacity: The key for the GoM to integrate drones
in the existing supply chain system, in the long run, is the issue
of locally based operation. To keep operations local, human
resources, including trained remote pilots, drone operators,
and technical project managers, are essential. Take BVLOS
for example: there is an increased risk in these applications
as human control is minimum, and the drones rely primarily
on computer systems [19], [20]. Specialized pilots with tech-
nical skills, including both knowledge and experience, such as
uploading flight plans in the case of an inconsistent mobile
network, are needed. These pilots are costly, and are typically
unwilling to stay for a long period of time, due to their unfamil-
iarity with the challenging operational environments. Although
Malawi has some capacity to fly drones, the GoM has not yet
been able to systematically develop such capacity. In the con-
text of this project, there was an agreement between the GoM
and the operators that, once the testing proves to be success-
ful, they will have Malawian pilots trained locally in order to
ensure the sustainability of the project. This, unfortunately, had
not yet taken place, as the project was still in the early phase
at the time of our study. Battery charging and cargo loading
offer two more examples. Health personnel may have a role
in loading/unloading a drone, confirming schedules, securing
a landing or dropping site, documenting deliveries, and even
launching the drone—each would require training and time [8].
If drone operations were to be scaled in Malawi, existing health
personnel need to be retrained with these responsibilities. In
Authorized licensed use limited to: MAIN LIBRARY UNIVERSITY OF ZURICH. Downloaded on March 27,2021 at 02:49:49 UTC from IEEE Xplore. Restrictions apply.
WANG: “AS IT IS AFRICA, IT IS OK”? ETHICAL CONSIDERATIONS OF DEVELOPMENT USE OF DRONES FOR DELIVERY IN MALAWI 27
reality, however, the MoH saw that the donors came in with
their own operations teams, and few health personnel have
been trained so far. It remains unclear whether, and if so when,
the GoM will be able to build the local capacity to manage the
technical processes independently—how many batteries they
will need to manage, who will handle the technical problems
associated with loading cargos in drones, and how they will
organize the maintenance of the drones in case of mechanical
errors or crashes, etc.
The drone operation is a combination of the technicalities
of the drone and project management knowledge, includ-
ing especially the implementation of the drone operation in
resource-constrained environments. The management of drone
operations involves complex planning, including landscape
research, identifying the country, designing the activity, suc-
cessful procurement, contracting, stakeholder management,
and getting approvals for flights. Additional aspects include
hiring local staff, getting buy-in from local communities, sen-
sitizing local communities, creating communication processes
with medical staff for deliveries and pick up, as well as over-
all management of the budget and subcontractors. Similarly,
testing, licensing, and certification of drones need to be done
and need to be streamlined and made local and affordable.
In countries where technical resources and human capacity
are limited, such as in Malawi, an international drone service
provider may present a valuable option during an initial phase.
Ultimately, local capacity needs to be strengthened as drone
service providers can be costly, and might not always be able
to deliver the optimal solution for every setting [25]. For sus-
tainable in-country drone operation and its maintenance in the
long run, local capacity building is a critical factor and should
be demanded by governments, facilitated by implementers,
and supported by donors [26]. As with any innovative health
intervention, the sustainability of drone-supported healthcare
systems will further necessitate strong capacity building, an
efficient impact monitoring and evaluation cycle, as well as
in-country commitment, including investment in drone reg-
ulations, project design, and long-term ownership [5]. This
involves multistakeholder consultation, and potentially the
development of new regulations which, again, implies costs
as well as the need for local capacity.
With respect to airspace management, the touchy issue
of drone regulation is at play. Air space is highly regu-
lated by civil and international aviation authorities, especially
with respect to manned aircraft where air space regulation
is currently much more established than in the unmanned
space. Although drones fly at much lower altitudes than
most manned aircraft, all aircraft pass through low-altitude
space, thereby requiring coordination. For drones, techni-
cal issues relate to battery life, payload capacity, and the
ability to detect and avoid any problems from the control
station. Safety and security issues include the drone’s abil-
ity to avoid near misses, collisions, and accidents as well
as hijacking and espionage [19], [20]. Administrative issues
related to drone operator training and licensing and the service
provider’s legal compliance, fiscal health and compliance, and
service costs [19], [20]. To achieve continued growth in drone
applications, drones must meet or exceed the requirements
specified in each of these regulatory areas [19]. Regulators
should therefore think beyond borders and in the longer-term
about the need for integrating drones in the existing logistics
systems in general [5], [8]. Like many countries in the world,
Malawi did not have drone-specific regulations at the onset.
Through a lengthy process of learning, Malawi started its own
regulation development in 2016, and now has drone pilots who
are not only certified to fly drones, but are involved in drafting
national drone regulations. However, as the main focus at the
DCA is manned aircrafts, drones create an additional job for
the staff, requiring dedicated manpower to supervise. In civil
aviation, there is undefined global authority—ICAO acts as
a facilitator and coordinator, while governments expect ICAO
to take a lead with standards, following what they have done
with manned aircrafts. However, due to a lack of prior experi-
ence with drones, ICAO is also learning from the industry and
seeking help with setting up the appropriate procedures glob-
ally. In the long run, airspace governance for drones will need
to address the interlinked issues of safety, privacy, account-
ability, and sovereignty. To this end, ICAO and governments
will have to anticipate future uses of the drone technology,
including humanitarian, development, and healthcare applica-
tions, taking into consideration the interests of a broad array
of stakeholders [1].
2) Donor Dependence: For countries like Malawi, govern-
ments need committed resources, both human and financial,
to improve infrastructure and provide services to their citi-
zens. Traditionally, if the GoM identifies needs and gaps, they
will approach donors and solicit funds, donors will then evalu-
ate their “aid-worthiness” and allocate funds accordingly. This
leaves the GoM in a conundrum: for donor-funded projects,
the donors direct what they want to do with their funds, decide
which issue or location to tackle, and whether to commit fur-
ther funds or withdraw—all of which are beyond the reach
of the GoM. Even when donors partner with the MoH, such
as in this project, the MoH is not in the position to nego-
tiate possibilities regarding how to allocate resources. If the
donor chooses HIV, then malaria is out of the scope, even
if there may be bigger needs [27]. And if the donor funds
run out, then the project ends, even if the health outcome
would be setback and those that are involved would be neg-
atively impacted. In the case of this project, unlike mainland
Malawi where land-based health supply chains are available,
interruption of donor funds would be especially harsh for the
villagers on the islands, and the seemingly improved patient
care may well become a once-in-a-life-time experience. Take
cargo drones for example: the payload of a drone must make
not only technical but also economic sense; otherwise, flying
a drone would not be dissimilar to flying a manned aircraft. As
the drone technology has evolved, challenges have arisen as
regards aligning technology partners with stakeholder needs.
How to ensure that drone applications will address relevant
problems by teams who understand both the technicalities of
the technology and the local contexts remains a challenge [8].
The rise of medical cargo drones in Malawi has its structural
roots because the existing problems are multifaceted—some
are location targeted, others are issue targeted; consequently,
some districts have a lot of donor supports, while others are
Authorized licensed use limited to: MAIN LIBRARY UNIVERSITY OF ZURICH. Downloaded on March 27,2021 at 02:49:49 UTC from IEEE Xplore. Restrictions apply.
28 IEEE TRANSACTIONS ON TECHNOLOGY AND SOCIETY, VOL. 2, NO. 1, MARCH 2021
suffering. Malaria, for instance, is an issue across sub-Saharan
Africa; HIV has been prevalent since the 1980s till today and
is bound to stay. According to the interview participants, the
health sector of Malawi receives a lot of innovation initiatives
from donors, and as long as they help improve the health
system and potentially save lives, the MoH would support
them. At the same time, there has been an outcry among the
District Health Officers that, sometimes, the donors “want to
do what they want and not what we want.” The ideal way of
addressing the health challenges of Malawi should be district
mapping according to respective health needs, driven by the
MoH. In reality, however, when donors come to Malawi, they
may come with their own ideas regardless of the particular
conditions and capacity at the local level. Since it is addi-
tional resources which will assist in one way or another, the
GoM seizes the funding opportunity. And when those initia-
tives are to be implemented, the decisions may have already
been made at, and passed through, the central level. Districts
may not be informed prior, or consulted about which donors
are needed, and which areas are to be supported. Given its
poor infrastructure and a lack of local capacity, to what extent
drones can really offer a solution to tackle these intersec-
tional and complex systemic problems of Malawi remains
to be seen.
One reason leading to donor hesitance in committing to
long-term investment in medical cargo drones is the lack of
knowledge about the outcomes of various pilot projects. The
health supply chain challenges of Malawi are cross-cutting
issues. Donors across the globe look at what health impacts
drones could really make, and whether local health capacity
may actually be strengthened by introducing drones, within the
time frame of operations [24]. According to the interviewed
MoH personnel, since not even one project has made signifi-
cant impact in Malawi so far, perceptions of donors regarding
what drones are capable of, and whether drone operations are
indeed “aid-worthy,” have been negatively affected. In terms
of the scalability of drones in countries like Malawi, it will be
dependent on a few factors: 1) the availability of local skills,
which is essential because the current experience shows that
maintaining the international staff in Malawi is neither fea-
sible nor cost effective; 2) the evidence of value creation of
adding drones to the existing health supply chain systems,
which requires a solid monitoring and evaluation framework
that could track progress; and 3) the maturity and cost of the
technology itself, as currently there are important drone parts
that cannot be manufactured locally and have to be imported,
the purchasing, maintaining, and repairing of the drone can
thus be burdensome. And finally, the question about the busi-
ness sense in manufacturing, which depends on the use case
and whether there is a business opportunity. The global drone
community has a culture of optimism—people see a few tests
and assume that everything is figured out; but there is actu-
ally still a lot more to understand. Until these questions are
answered, unless there is some donor who firmly believes
in the technology and decides to invest in massive scalabil-
ity regardless, the overoptimistic scenario where the operation
costs will get lower and local skilled labor becomes available
for running drones would be far-fetched.
IV. DISCUSSION
It is often perceived that the biggest international develop-
ment challenge is the aid sector itself, as everything is based
on donors; yet the interconnectedness among, and the com-
plexity of, human development issues are not always fully
recognized by donors working in the development space [27].
Beneficiaries would expect that if donors are providing a ser-
vice, they would continue providing it without interruptions.
Such a perception commonly exists in projects involving donor
funds, which can be decreased or discontinued. There is,
hence, the risk that expectation and dependence are created,
while services fail to deliver due to operational reasons and
donor preferences and their funding behavior [27]. Reality
requires development actors creating a system that benefits
populations in need, while keeping in mind that projects are
determined by funding and may come to an end. The fact
that donor funds may exhaust, and projects may terminate—
sometimes after people have got used to the benefits—may not
be an ethical issue in itself, but it is a limitation that the devel-
opment sector faces historically [27], [28]. In the case of med-
ical cargo drones, the discussions have been that if the drones
are used to bring medical supplies to populations in need, then
such innovation helps generate greater public good. Traditional
means of transportation do not involve innovation, but they
can also create dependence, can cause potential risks, and can
fail to deliver. Therefore, giving a chance to demonstrate the
potentials of drones as an alternative to the existing health
logistics does add value. This may be true to some degree;
the danger, however, is that in the development sector, people
often look for shortcuts and drones provide such an option [9].
The least-developed countries across the world have the same
challenge of lack of development of infrastructure. In Africa,
the biggest challenge tends to be roads, which fuels the nar-
rative that “Africa needs more drones than roads” [9]. As this
case study illustrates, Malawi has its unique environments that
other countries do not have, some of which are in favor of
using drones (e.g., government support and health need), oth-
ers are disadvantageous (e.g., mobile connectivity and weather
conditions). This raises the question as to whether drones
indeed offer an appropriate solution for countries like Malawi
in the first place, which currently has no existing guidance
for decision makers to refer to when developing innovation
strategies.
With respect to public acceptability of innovation, there
are three levels to it: 1) the donors—how much they truly
believe in the technology, and what tradeoffs they perceive;
2) the local government—how risk averse they are, and how
they could benefit based on rational calculations; and 3) the
community—how much they understand new technology, and
how they associate it with local beliefs, norms and values.
For donors, when a new technology is introduced, they assess
whether it offers solutions to existing problems, how best to
utilize it as a tool to benefit public health or development, what
makes sense, and what is a waste of the resources. For local
governments, although there are issues around resource avail-
ability and local capacity, innovation is generally welcomed,
as long as it makes positive change and visible impacts. For
Authorized licensed use limited to: MAIN LIBRARY UNIVERSITY OF ZURICH. Downloaded on March 27,2021 at 02:49:49 UTC from IEEE Xplore. Restrictions apply.
WANG: “AS IT IS AFRICA, IT IS OK”? ETHICAL CONSIDERATIONS OF DEVELOPMENT USE OF DRONES FOR DELIVERY IN MALAWI 29
communities, they need to be actively engaged in the process,
starting from an accurate understanding about the potential
risks of the technology to be introduced. In terms of drones,
as there is not much experience anywhere in the world, the
business model for drones is not yet clear [5], [8]. Is it better
to outsource drone services to the private sector, or to build
in-house capacity within implementing organizations? What
sort of monitoring and evaluation systems are needed? These
questions call for practical guidance to facilitate decision mak-
ing. Although the indicators of success are unlikely to change
with the introduction of drones, measuring how performance
changes around the introduction of this new technology in
particular settings will be key to understand its value [8]. As
technology evolves, the general public’s understanding of how
things work evolves, and the government’s tolerance and will-
ingness to give approvals evolves as well. Just as with anything
else, the more experience and exposure people have to new
technology, the more familiarity with it they develop, and the
better equipped they are to understand the nuances, the sug-
gested benefits, as well as the potential risks—which may or
may not cause actual problems, but the awareness of them is
crucial. Although drone delivery is likely to advance quickly
over the next few years, it remains unclear whether, outside
of limited delivery of small payload of medical supplies, there
is a clear case for drones to be deployed specifically to assist
international aid [10].
Regarding the technology experimentation aspect, a lot of
innovation initiatives have been proposed to Malawi in the
name of study or testing. Some were rejected by the GoM,
others were pushed on and got implemented in different dis-
tricts. In our study, interviewees from the health sector raised
questions around why “testing things out on Malawians before
giving it to others,” why accepting it when “they make their
need our need,” and why not insisting that “we want something
that we want.” While acknowledging the importance of foreign
aid, they were disappointed about “feeling powerless because
we are poor,” and wished to have the “do not impose but ask”
approach. For decades, Malawi has been a relatively “free”
space—not just for drones but for development projects in
general [27], [28]. Countries like Malawi are receptive to aid
because there are needs and gaps everywhere, rendering them
particularly vulnerable to “solutions,” including those that are
looking for a problem. The involvement of the tech sector in
the development world may not necessarily be negative, but
the emphasis should be on a careful and holistic approach,
such that beneficiary countries can be truly empowered by
the proposed innovation, and can actually manage its appli-
cations in their own capacity. Government institutions play
a critical role in facilitating project approval, negotiating with
regulatory bodies and ministries, and coordinating a country
approach among all health stakeholders even in cases where
domestic funding cannot be provided [5]. After all, drones can
be disruptive in both good and bad ways. In a positive sense,
their introduction can potentially break down existing barri-
ers between departments who previously worked in silos, and
help integrate the health supply chain systems through coor-
dinated efforts. The tension lies between insider and outsider
tactics in the use of drones—policymakers and innovators alike
should engage in a broad and inclusive discussion about how
principles might be best balanced, as public accountability is
humanitarianism’s corner stone [8]. Ultimately, aid provision
is not about drones, but about being able to serve populations
in need. Challenges arise as to how best to engage the pri-
vate sector, which lacks well-thought-through guidance in its
current state of development.
The drone industry expansion from the civilian space to
the humanitarian and development world, where drones cre-
ate a significant new market while potentially adding value
such as delivering medicines, has been aggressive in recent
years [10], [29]. As the industry eyes civilian applications for
its products, it is likely that governments, as well as aid agen-
cies, will be subject to extensive lobbying efforts, including
both the procurement of the so-called “humanitarian drones”
and the push for the inclusion of it as part of international
engagements [30]. Humanitarian and development organiza-
tions must, therefore, carefully consider the practical, ethical,
and legal implications of these developments. In terms of
tradeoffs, would the cost of the drones divert resources from
a better use? Do drones represent a risk to affected populations,
which is at least as great as the risk to be eliminated by the
health sector? Will drones remove work from, or create addi-
tional work to, people who are currently in the workforce?
How would drones, as well as the aid organizations behind
them, be perceived by locals who may have hardly ever seen,
or may even have negative experiences with, flying object?
How would they experience routine flights over their village
or their houses and heads? Would it be something culturally
acceptable or accepted with hostility? How will the drone
industry contribute to building the interdisciplinary capacity as
many seemingly unrelated sectors could actually be affected?
These questions, again, call for actionable guidance, which
needs to be evaluated against a comprehensive needs assess-
ment regarding existing resources and bottlenecks in the health
supply chain system and associated issue areas. The com-
plexity of health logistics in low-resource settings requires
dedicated stakeholder engagement, and the drone industry is
an integral part of it. This is not only the responsibility of
a government but also the industry should make strides in the
analysis of demand and the network of the issues of concerns
as well—before proposing and initiating a drone project.
V. CONCLUSION
While the defense and intelligence sectors were the early
adopters of drones, the development sector, along with the
commercial sector, has begun to leverage the technology to
reach the hard-to-reach. To counteract the negative percep-
tion of drones and avoid the more stringent regulations in the
Global North, Africa has become a hotspot for the drone indus-
try to develop technology and to obtain legitimacy [9], [31].
Most development programs utilizing drones in the health
sector aim to supplement the traditional health logistics in
geographically challenging areas and low-resource environ-
ments, through collaboration with governments, the private
sector, and international organizations [30]. As the drone tech-
nology advances, it is likely that the use of drones for
logistics in global health will become increasingly viable
as—when deployed responsibly and effectively—the drone
Authorized licensed use limited to: MAIN LIBRARY UNIVERSITY OF ZURICH. Downloaded on March 27,2021 at 02:49:49 UTC from IEEE Xplore. Restrictions apply.
30 IEEE TRANSACTIONS ON TECHNOLOGY AND SOCIETY, VOL. 2, NO. 1, MARCH 2021
technology can potentially add value to enhance health sup-
ply chain performance, improve access to health services,
and create greater public health outcomes. As the aid sector
embraces innovation, it opens up a unique space for technol-
ogy, which may be deployed with a “solutionist” mentality
by aid organizations [4], [11], [12]. Whether these innova-
tion proposals can substantively solve the actual problems
of the countries they intend to serve, in a sustainable man-
ner and in the long run, remains to be further examined
and elucidated. As proposed in this article, ultimately, the
key lies with ensuring rigorous reflections about the ethical
challenges technological innovation may invoke, developing
responsive methodologies to assess its potential for harms
relative to potential for benefits, and establishing actionable
ethical guidance to identify, address, and tackle such chal-
lenges. Following these insights, future work should strive to
establish a humanitarian innovation framework that is value-
sensitive and context specific, and geared toward determining
practical courses of action to address these concerns.
ACKNOWLEDGMENT
In developing the ideas presented in this article, the
author received insightful feedback from Dr. Markus Christen
(University of Zurich, Switzerland), and Dr. Matthew Hunt
(McGill University, Canada), to whom the author is immensely
indebted. The author is also grateful to Dr. Heather Collister
for assistance with the interview data, and local stakeholders
for fact validation, during the data analysis process. Her deep-
est gratitude goes to the project partners who supported her in
her fieldwork, without whom this case study would not have
been conceived in the first place.
REFERENCES
[1] The World Bank. Understanding Poverty. Accessed: Jan. 6, 2021.
[Online]. Available: https://www.worldbank.org/en/topic/poverty/
overview
[2] J. C. Chow, The Case for Humanitarian Drones, OpenCanada, Toronto,
ON, Canada, 2012. Accessed: Jan. 6, 2021. [Online]. Available: https:
//www.opencanada.org/features/the-casefor-humanitarian-drones/
[3] Foundation for Mine Action. (2016). Drones in Humanitarian Action:
A Guide to the Use of Airborne Systems in Humanitarian Crises.
Accessed: Jan. 6, 2021. [Online]. Available: http://drones.fsd.ch/wp-
content/uploads/2016/11/Drones-in-Humanitarian-Action.pdf
[4] N. Wang, “‘Killing two birds with one stone’? A case study of devel-
opment use of drones,” in Proc. IEEE Int. Symp. Technol. Soc. (ISTAS),
2021, pp. 1–8.
[5] A. M. Knoblauch et al., “Bi-directional drones to strengthen health-
care provision: Experiences and lessons from Madagascar, Malawi and
Senegal,” BMJ Global Health, vol. 4, no. 4, pp. 1–10, 2019. [Online].
Available: https://www.doi.org/10.1136/bmjgh-2019-001541
[6] C. Howell et al., “The first government sanctioned delivery of medical
supplies by remotely controlled unmanned aerial system (UAS),” NASA,
Washington, DC, USA, Rep. NF1676L-23651, 2016. Accessed: Jan. 6,
2021. [Online]. Available: https://ntrs.nasa.gov/citations/20160009120
[7] Zipline. Global Public Health. Accessed: Jan. 6, 2021. [Online].
Available: https://flyzipline.com/solutions/global-public-health/
[8] M. Eichleay, E. Evens, K. Stankevitz, and C. Parker, “Using the
unmanned aerial vehicle delivery decision tool to consider transporting
medical supplies via drone,” Global Health Sci. Pract. J., vol. 7, no. 4,
pp. 500–506, 2019. [Online]. Available: https://doi.org/10.9745/GHSP-
D-19-00119
[9] K. B. Sandvik, “African drone stories,” Behemoth J.
Civilisation, vol. 8, no. 2, pp. 73–96, 2015. [Online]. Available:
https://doi.org/10.6094/behemoth.2015.8.2.870
[10] N. A. Raymond, B. Card, Z. Al Achkar, The Case Against
Humanitarian Drones, OpenCanada, Toronto, ON, Canada, 2012.
Accessed: Jan. 6, 2021. [Online]. Available: https://www.opencanada.
org/features/the-case-against-humanitarian-drones/
[11] N. Wang, “‘A successful story that can be sold’? A case
study of humanitarian use of drones,” in Proc. IEEE Int. Symp.
Technol. Soci. (ISTAS), 2019, pp. 1–6. [Online]. Available: https://
doi.org/10.1109/ISTAS48451.2019.8938015
[12] N. Wang, “‘We live on hope...’: Ethical considerations of human-
itarian use of drones in post-disaster nepal,” IEEE Technol. Soc.
Mag., vol. 39, no. 3, pp. 76–85, Sep. 2020. [Online]. Available:
https://doi.org/10.1109/MTS.2020.3012332
[13] Y. Lincoln and E. Guba, “Paradigmatic controversies, contradictions, and
emerging confluences, revisited,” in The Sage Handbook of Qualitative
Research, 3rd ed., N. K. Denzin, and Y. S. Lincoln, Eds. Thousand Oaks,
CA, USA: SAGE, 2011, pp. 191–215.
[14] R. Yin, Case Study Research: Design and Methods, 5th ed. Los Angeles,
CA, USA: SAGE, 2013.
[15] M. Sandelowski, “Whatever happened to qualitative description?” Res.
Nurs. Health, vol. 23, no. 4, pp. 334–340, 2000.
[16] M. A. Neergaard, F. Olesen, R. S. Andersen, and J. Sondergaard,
“Qualitative description—The poor cousin of health research?” BMC
Med. Res. Methodol., vol. 9, p. 52, Jul. 2009.
[17] USAID Global Health Supply Chain Program-Procurement and
Supply Management, Unmanned Aerial Vehicles Landscape Analysis:
Applications in the Development Context, Chemonics Int., Inc.,
Washington, DC, USA, 2017. Accessed: Jan. 6, 2021. [Online].
Available: https://www.ghsupplychain.org/sites/default/files/2017-06/
GHSC_PSM_UAV%20Analysis_final.pdf
[18] P. Meier, Digital Humanitarians: How Big Data Is Changing the Face
of Humanitarian Response. New York, NY, USA: Routledge, 2015, ch.
4. [Online]. Available: https://doi.org/10.1201/b18023
[19] M. Balasingam, “Drones in medicine—The rise of the machines,” Int. J.
Clin. Pract., vol. 71, no. 9, 2017, Art. no. e12989. [Online]. Available:
https://doi.org/10.1111/ijcp.12989
[20] K. Bhatt, A. Pourmand, and N. Sikka, “Targeted applications
of unmanned aerial vehicles (drones) in telemedicine,” Telemed.
e-Health, vol. 24, no. 11, pp. 833–838, 2018. [Online]. Available:
https://doi.org/10.1089/tmj.2017.0289
[21] D. Cawthorne and A. Robbins-Van Wynsberghe, “From healthdrone to
frugaldrone: Value-sensitive design of a blood sample transportation
drone,” in Proc. IEEE Int. Symp. Technol. Soc. (ISTAS), 2019, pp. 1–7.
[Online]. Available: https://doi.org/10.1109/ISTAS48451.2019.8938019
[22] Unmanned Aircraft Systems (UAS) For Humanitarian Aid and
Emergency, Int. Civil Aviation Org. (ICAO), Montreal, QC,
Canada, 2020. Accessed: Jan. 6, 2021. [Online]. Available https:
//www.icao.int/safety/UA/UAID/Documents/ICAO%20UAID%
20Guidance%20Material.pdf
[23] The Logistics Handbook: A Practical Guide for the Supply
Chain Management of Health Commodities, United States Agency
Int. Develop. (USAID), Washington, DC, USA, 2011. Accessed:
Jan. 6, 2021. [Online]. Available: https://www.ghsupplychain.org
/sites/default/files/2019-07/Logistics%20Handbook.pdf
[24] UAV for Payload Delivery Working Group (UPDWG). Market
Considerations to Accelerate Medical Drone Delivery at Scale.
Accessed: Jan. 6, 2021. [Online]. Available: https://www.youtube.com
/watch?mc_cid=4a0bd5bafd&mc_eid=0ae675bf84&v=nW4lGklI_vg&
feature=youtu.be
[25] W. Dirks, “Evaluation of the business case for cargo drones in human-
itarian action: Qualitative analysis,” Res. Develop. Dept., Médecins
Sans Frontières (MSF), Tokyo, Japan, Rep., 2017. Accessed: Jan.
6, 2021. [Online]. Available: https://s3-eu-west-1.amazonaws.com/
s3.sourceafrica.net/documents/120074/Drones.pdf
[26] N. Phillips et al.,Costs Associated With the Use of Unmanned
Aerial Vehicles for Transportation of Laboratory Samples in Malawi,
VillageReach, Seattle, WA, USA, 2016. Accessed on: Jan. 6, 2021.
[Online]. Available: http://www.villagereach.org/wp-content/uploads/
2017/06/Malawi-UAS-Report_MOH-Draft_-FINAL_14_07_16.pdf
[27] N. S. Morfit, “‘AIDS is money’ How donor preferences reconfigure local
realities,” World Develop., vol. 39, no. 1, pp. 64–76, 2011. [Online].
Available: https://doi.org/10.1016/j.worlddev.2010.09.003
[28] J. N. Brass, W. Longhofer, R. S. Robinson, and A. Schnable, “NGOs and
international development: A review of thirty-five years of scholarship,”
World Develop., vol. 112, pp. 136–149, Dec. 2018. [Online]. Available:
https://doi.org/10.1016/j.worlddev.2018.07.016
[29] K. B. Sandvik and K. Lohne, “The rise of the humanitarian
drone: Giving content to an emerging concept,” Millennium J. Int.
Studies, vol. 43, no. 1, pp. 145–164, 2014. [Online]. Available:
https://doi.org/10.1177/0305829814529470
[30] K. B. Sandvik and K. Lohne, The Promise and Perils of ‘Disaster
Drones’. London, U.K.: Humanitarian Pract. Netw., 2013. Accessed: Jan.
6, 2021. [Online]. Available: https://odihpn.org/magazine/the-promise-
and-perils-of-%C2%91disasterdrones%C2%92/
[31] Drones for Health: Boosting Access to Life Saving Products,
VillageReach, Seattle, WA, USA, 2019. Accessed: Jan. 6, 2021.
[Online]. Available: https://www.villagereach.org/wp-content/uploads/
2019/09/VillageReach_DronesForHealth_Overview.pdf
Authorized licensed use limited to: MAIN LIBRARY UNIVERSITY OF ZURICH. Downloaded on March 27,2021 at 02:49:49 UTC from IEEE Xplore. Restrictions apply.
