Cargo Drones in Humanitarian Contexts Meeting Summary

Introduction

With funding from DG ECHO, FSD is implementing an initiative on Drones in Humanitarian Action in collaboration with CartONG, Zoï Environment Network and the Humanitarian UAV Network (UAViators)[1]. Discussions among the partners and participants led to a meeting focusing on cargo drone applications in humanitarian contexts co-organized and hosted by the University of Sheffield on 20-21 June 2016. The meeting brought together the principle stakeholders interested in the use of cargo drones for the delivery of essential humanitarian payloads. The purpose of this meeting was to advance the effective use of this emerging technology in humanitarian efforts actively, responsibly and effectively. This document summarizes the presentations and main discussion items of this meeting.

PDF Summary Report

Background

The 24 participants who contributed to the discussions included representatives of humanitarian NGOs and businesses, as well as technology experts and donors. The outcomes of this discussion will provide direct input for guidelines on the use of cargo drones in humanitarian action, to be published later in 2016.

Agenda Summary

The consultation was conducted over two days and comprised a keynote address by the University of Sheffield and select presentations on Humanitarian Needs, Logistical/CBA Needs and Technological Needs. In addition, small breakout discussion groups were organized on the topics of Regulations; Technology; Sustainability and Business Models/Big Data; Realizing Supply Chain Improvements; and Code of Conduct. Other experts focused on cross-cutting themes and presented their viewpoints during an informative round of lightning talks.

Presentation Summaries

The full agenda of the Experts Meeting is available in Annex I along with links to the slides used by each speaker. This section briefly summarizes the main points provided in each of the formal presentations.

Professor Lenny Koh (Presentation Slides) from the University of Sheffield’s Logistics and Supply Chain Management Research Centre and Advanced Resource Efficiency Centre opened the Experts Meeting with a keynote presentation focused on the future of cargo logistics platforms. She took a holistic approach by broadening her scope beyond aerial robotics (drones) to autonomous robotics systems in general. She subsequently brought attention to a number of important peer-reviewed academic papers including:

  • Özdamar, L., & Ertem, M. A. (2015). Models, solutions and enabling technologies in humanitarian logistics. European Journal of Operational Research, 244(1), 55-65.
  • Erdelj, M., & Natalizio, E. (2016, February). UAV-assisted disaster management: Applications and open issues. In 2016 International Conference on Computing, Networking and Communications (ICNC) (pp. 1-5).

Professor Koh also homed in on key themes such as security and regulations, optimizing the supply chain through robotics and the technology and human needs that are essential for cargo robotics. She noted that over 1 billion people around the world live in remote and difficult to reach communities without regular access to medicine. What’s more, inadequate last-mile delivery logistics are the critical constraint preventing medical supplies from reaching these areas, and cause 40 per cent of vaccines supplied to parts of the underdeveloped world to expire before they can be administered. Last-mile delivery is usually affected by such barriers as island isolation; the breakdown or lack of a basic transport system and connectivity; the lack of secure storage and warehouse facilities; and an inefficient distribution process and supply chain resulting in high wastage. In closing, Professor Koh emphasized the need to get out of the lab and controlled testing environments in order to properly field-test new solutions.

Denise Soesilo (Presentation Slides) from FSD introduced the DG ECHO-funded initiative on Drones in Humanitarian Action. The purpose of the initiative – spearheaded by FSD with partners – is to consolidate existing knowledge about the use of drones in the humanitarian context, and to further test, promote and disseminate their appropriate use. This meeting on cargo drones was convened to understand the current state of the technology and to serve as a baseline for future work. In addition, the participants were asked to work together to identify the best use cases of cargo drones in humanitarian contexts and the main challenges to safe and effective implementation. In closing, Denise explained that a summary of this meeting would be shared with the public following a review from the participants. The program’s final deliverable will be a major report combining insights and lessons learned from multiple consultation meetings, case studies and field-tests. This final report will be shared with DG ECHO, humanitarian actors and all other stakeholders to ensure that UAVs can be fully leveraged in humanitarian efforts.

Dr. Patrick Meier (Presentation Slides) from WeRobotics invited participants to keep the bigger picture in mind vis-à-vis cargo drones in humanitarian action. He framed the meeting within the broader context of robotics and also within the context of the World Humanitarian Summit (WHS). He noted that the international community spends about US $25 billion per year to provide life-saving assistance to people devastated by wars and natural disasters. Citing the United Nations, Patrick noted that this figure is US $15 billion short of what is urgently needed per year. He suggests that one way to deal with this massive gap in funding is to radically increase the efficiency and productivity of the humanitarian industry. This, he explained, is where Artificial Intelligence and Autonomous Robotics come in. The World Economic Forum calls this powerful combination the Fourth Industrial Revolution. And whether humanitarian organizations like it or not, this revolution will impact the aid industry just like it is already impacting other sectors through radical gains in productivity and efficiency. Patrick notes that the heart of this revolution is the shift from manually controlled technology to increasingly intelligent and autonomous systems. As such, he echoed Professor Koh’s keynote by broadening the discussion beyond aerial robotics to include maritime and terrestrial robotics as well.

This shift towards AI-powered robotics has direct implications for both data collection and cargo transportation in the humanitarian space. But he warned that these implications must not be imposed top-down. This is where the WHS comes in. As Patrick noted, one of the main messages to come out of the summit was the importance of localizing aid. “So how do we localize the appropriate use of robotics?” he asked. In closing, Patrick emphasized the need for a reality check as far as the technology behind cargo robotics goes. Demonstrations of this new technology in perfect weather conditions do not qualify as rigorous field tests or adequate proof of their reliability. The engineering challenge is far more difficult, costly and time-consuming than most companies in this space realize. To this end, rigorous field testing is key, which, when combined with the need to localize this technology, requires a code of conduct to inform the safe, responsible and effective use of cargo UAVs in humanitarian settings.

Daniel Ronen from the Last Hundred Miles Research Group and the company UAVaid spoke about cargo drones as enablers of supply chain optimization. One third of the world’s population lacks regular access to essential medicines. The causes include poor infrastructure in developing countries, complex logistics and lack of money. In sub-Saharan Africa, for example, about 40 per cent of vaccines expire before the final leg of transportation. What’s more, traditional last-mile delivery methods are push-based and thus often unreliable since this approach is driven by predicted rather than actual demand. Forecast-based ordering requires data that are often missing or incorrect. High inventory levels are used to compensate for the resulting variability and unreliability. This in turn can lead to high wastage levels. These logistical problems are further exacerbated by cold chain requirements. Almost all vaccines and antivenin are thermo sensitive and thus need to be stored within a temperature band. Daniel then highlighted the advantages of pull-based methods in contrast to push-based supply chain methods. The former has lower inventory requirements and lower wastage while offering higher responsiveness to changes in actual demand (during emergencies, for example). In reality, Daniel notes, supply chains are often made up of hybrid push-pull systems with the question being where to use push and where to lead with pull instead.

Daniel used the second half of his presentation to explain the role that drones can play in final mile delivery. Rural health workers can contact district stores with exact requirements based on actual demands. In return, the drones can deliver exactly what is needed with no supply guesswork required. Delivery can be carried out in minutes rather than hours or days irrespective of ground conditions or infrastructure. This can lead to more reliable and consistent delivery as well as multiple deliveries per day. As a result, the delivery system becomes responsive and no longer presents a logistical constraint. With actual real-time demand data, each node in the system will theoretically be optimized individually thus requiring minimal inventory levels and greatly reducing wastage due to the overstock of medications or the delivery of the wrong medications. In sum, drones can reduce the uncertainty in the final mile and potentially eliminate final mile logistical constraints. They become an enabler for efficiencies along the supply chain, making the system responsive rather than predictive.

Joel Kaiser (Presentation Slides) from Medair spoke about the perceived challenges in the humanitarian use of cargo drones and highlighted the challenges that humanitarian organizations face when considering the use of cargo drones. But first, Joel made a distinction about the difference between humanitarian aid and development aid. Humanitarian aid is usually short-term help aimed at saving lives, reducing suffering and respecting human dignity by providing material and logistic assistance to people in need. It may be distinguished from development aid, which seeks to address the underlying socioeconomic factors that may have led to a crisis or emergency. As such, development aid differs from humanitarian aid since the former focuses on alleviating poverty in the long term.

Joel’s presentation focused specifically on the challenges that humanitarian organizations perceive when considering the use of cargo drones in the context of humanitarian aid. More specifically, Joel highlighted four key challenges – Capital, Community, Capacity and Conflict. Capital: Non-profits struggle to invest in proven technology. The rate of change in civilian drone technology is rapid and the technology with respect to cargo drones is still nascent and largely unproven. To this end, what financing models can lessen the risk when investing in these new technologies? Community: Drones aiding affected communities can do a lot of good. Projects that lack local support, however, are simply unsustainable. How do we involve affected communities? Capacity: High-value/low-volume cargo is easier to identify in terms of use-cases. This means that cargo may be more beneficial as a contingency, not a certainty. How do we invert this? Conflict: Humanitarian crises are volatile. Contingencies are the norm. Drones could reach areas that aid workers cannot. What is the most prudent way to pursue this in the context of conflict zones? In conclusion, Joel explained the relevance of logistical efforts across a spectrum of disaster categories and attributes, noting that logistical efforts are most relevant in sudden-onset natural disasters as opposed to man-made or slow-onset crises.

Rae McGrath (Presentation Slides) from Mercy Corps spoke on the use of UAVs and humanitarian aid in modern conflict. Asymmetrical conflict is increasingly the working environment for agencies involved in humanitarian assistance to civilian communities. Syria has offered the most intense example during the past five years of how providing assistance in a conflict involving multiple combatant forces with fast shifting frontlines requires that responding agencies must either adapt their response modalities to match the battlefield or become irrelevant. The reality is that we have developed, out of necessity, a system of asymmetrical aid – a blunt recognition that, in the new warfare environment, we are not and cannot be, observers. The battlefield is our office, and we are, albeit non-violently and without control, inextricably part of the conflict. So, in such circumstances how could we make use of unarmed, civilian UAVs in future conflicts?

The negatives are obvious – cost and risk to staff. For starters, UAVs would need to be cost-efficient. But UAVs are also used by combatants and international observers in asymmetrical conflicts such as in Syria and Yemen, and UAV use by humanitarian agencies would raise suspicions about the agencies’ impartiality. The deployment of UAVs by humanitarian agencies in conflict situations would involve substantial training and staffing, and so must be fully supported by the international donor community, but Rae notes that we cannot afford to dismiss the use of any technology that may provide us with the means to save lives and maintain humanitarian support to civilians trapped in conflict. The potential for how UAVs may be deployed lies in a flexible approach, and organizations like Mercy Corps must continue to develop a toolbox approach that offers solutions to multiple field challenges. The question that Rae want us to consider is whether we can define certain deployments as one-off limited drops – drugs to hospitals and isolated communities, or limited amounts of aid to small groups of internally displaced groups moving through otherwise inaccessible areas, for instance. Rae acknowledges that these are early days, but if we continue this discussion and look for potential applications as UAV technology develops, we will find synergy and opportunities where UAVs can be used in conflict response by humanitarian agencies.

Jonathan Ledgard (Presentation Slides) from Redline emphasized the importance of building the necessary ground infrastructure to support long-range, heavy-lift cargo drones. To this end, he introduced the concept of “droneports” to service the next generation of cargo drones. Droneports will include a dedicated runway to accommodate drones from various companies including Redline and Blueline Drones. The former, according to Jonathan, will have a 50 km range and 10 kg payload capacity while the latter will have a 100 km range and 100 kg payload. The droneports will be open to multiple drone companies just like airports are available to multiple airlines.

Artist rendering of the “droneports” as envisioned by Redline. Credit: Redline and Norman Foster Foundation

Artist rendering of the “droneports” as envisioned by Redline. Credit: Redline and Norman Foster Foundation

While the idea of droneports may sound futuristic to many in the humanitarian sector, Redline is committed to building the first of three droneports in Rwanda within the next 12 months. Jonathan’s goal is to have 44 per cent of the country’s population serviced by three droneports. These droneports will also accommodate workspaces for maintenance and prototyping. Jonathan noted that he is not interested in replacing the last mile of supply chain logistics with autonomous drones given the massive youth unemployment in Rwanda. Currently, the delivery of medicines and vaccines in Rwanda is in part carried out using motorbikes, which offers employment opportunities to Rwandans. Instead, Jonathan is looking to have the droneports service the country’s district hospitals rather than smaller rural clinics.

Luca Palermo (Presentation Slides) from the Air Transport Unit of the World Food Programme (WFP) discussed WFP’s current logistics challenges, and shed light on the reality from the field in the Democratic Republic of the Congo (DRC). More specifically, he focused on four key topics: 1) Facts and figures of current logistical operations; 2) Challenges in humanitarian logistics; 3) The need for innovation to find additional suppliers and aircraft types; and 4) The feasibility of using cargo drones in WFP operations. As background, Lucas first offered a quick overview of the United Nations Humanitarian Air Service, which provides passenger and light cargo air services on behalf of the humanitarian community. To service its short-term aviation needs, however, WFP draws on chartered aircraft for cargo space. These needs include the provision of airlifts and airdrops of food deliveries during emergency operations, among others.

WFP Aviation also provides dedicated air services to a number of external clients to support specific organizational needs. In 2015, WFP Aviation chartered aircraft for the United Nations High Commissioner for Refugees in DRC, Ethiopia and Tanzania, for example. Currently, WFP is using airdrops to deliver food to over 100 000 people affected by the protracted conflict in Syria. For this operation, Luca reports that the limited available human capital presents the main bottleneck for the food drops. Luca also provided concrete figures vis-à-vis the volume of cargo that WFP delivers. Following the devastating earthquakes in Nepal in 2015, WFP logged close to 2 500 hours of flight time to transport over 2.5 million kg of humanitarian cargo. Even then, Luca reports that there was a significant backlog of cargo that needed to be transported in Nepal and there were simply not enough helicopters.

Other typical challenges include the need for spare parts, customs clearances, aircraft clearances, access to fuel supply, the availability of crew accommodations and visas. Aging aircraft and aircraft availability also represent challenges for WFP. When an aircraft is no longer available to service a route due to technical problems, for example, this can cause serious disruptions in the supply chain since replacement aircraft are rarely available at short notice. In the DRC, supply chain disruptions sometimes occur when pilots refuse to fly certain routes due to security issues.

In order to use UAVs for payload delivery, the World Food Programme will need to develop an internal operational concept and standard operating procedures, and will need to include UAVs as part of the services provided internally and externally by WFP Aviation. Luca stressed that any use of cargo UAVs would have to ensure high accuracy of location delivery if parachutes are used and very high accuracy of flight paths especially when operating in complex airspace. In conclusion, Luca noted that (given the current range and payload capacity of UAVs today) while there are few instances where UAVs could add value in WFP logistical operations, such as servicing insecure areas that are within 100 km, being able to respond to those cases with UAVs could have an important positive impact on overall logistics efforts. Ideally, however, WFP would want to use UAVs with a range of 100-500 km and a payload capacity of 200-500 kg.

Oriol Lopez (Presentation Slides) from Médecins Sans Frontières (MSF) provided an overview of the logistical needs based on MSF operations. MSF works in some isolated places with difficult access where transport is expensive and complicated. The capacity to send some small goods using UAVs could therefore improve the organization’s capacity to save lives. In addition, the use of autonomous UAVs could enable MSF to become more efficient with their resources. While the technology behind cargo UAVs is still developing and presents some risk, he believes that UAVs could still offer some game-changing advantages.

Delivery drone in Papua New Guinea during an MSF project testing the delivery of samples with RPAS. Credit: MSF Paris

Delivery drone in Papua New Guinea during an MSF project testing the delivery of samples with RPAS. Credit: MSF Paris

Oriol described an MSF collaboration with Matternet in Papua New Guinea in 2014[2] where field testing focused on the collection of tuberculosis sputum samples. Findings from this pilot project suggested that multi-rotor platforms may not be the best fit for the distances necessary in Papua New Guinea. To this end, Oriol is looking to organize field tests by year’s end with two companies that have taken a hybrid approach – autonomous platforms that can take off and land vertically like helicopters and fly horizontally like an airplane. The two companies are Quantum Systems (see presentation summary below) and Vayu (which had to cancel at the last minute and therefore was unable to attend the Experts Meeting). Besides TB sputum and other lab samples, MSF is also interested in exploring the use of UAVs to transport blood, vaccines, antivenin and oxytocin. Oriol emphasized that in order to use UAVs for cargo transportation, the platforms need to be hybrid designs, be automatic and very simple to use, and have a 60 km minimum range and a 1 kg minimum payload capacity. In addition, the platforms must be easy to set up and operate, and highly reliable while being easy to maintain.

In conclusion, Oriol emphasized the limitations of current regulatory parameters such as the need to operate by visual line of sight. This typical rule represents the most pressing regulatory barrier as it impedes the full added value that UAVs can bring to humanitarian efforts.

Florian Schneider (Presentation Slides) from Quantum Systems gave the final presentation of the first day. Quantum Systems (QS) is a German company (start-up) that manufactures hybrid UAVs. Florian explained the reasons his company decided to build hybrid solutions rather than multi-rotor or fixed-wing platforms – the former has limited flight speed and flight time while the latter is typically more difficult to operate. In addition, fixed-wing platforms used for cargo transportation often require a dedicated runway or a catapult to launch the UAVs. In contrast, hybrid UAVs perform well in terms of flight speed, flight time, ease of handling and vertical take-off and landing (VTOL). QS will launch the Quantum TRON VTOL UAV, which has a range of 150 km and a payload capacity of 2.5 kg in 2017. In 2019 QS plans to introduce a larger version, the Quantum VRT. It will be able to cover distances up to 250 km with 5 kg capacity.

QS has already field tested their TRON platform in difficult environments (high winds and 40+ degrees Celsius in the Dubai desert). In addition, QS recently partnered with the German Coast Guard to field test landing the TRON on a Coast Guard ship off the coast of Germany. The cost of the TRON is about €60 000. Renting the TRON is expected to become an option through third parties.

TRON

Quantum System’s TRON RPAS. Credit: Quantum Systems

On Day 2, select speakers were invited to make presentations on cross-cutting initiatives. A brief summary of each of the five presentations follows.

Sid Rupani (Presentation Slides) from LLamasoft introduced his company, which specializes in designing better supply chains. The company’s Global Impact Team applies cutting-edge supply chain technology and thought leadership to the world’s most important supply chain challenges. The Team customizes LLamasoft’s platform for individual clients and operates in 17 countries across Africa, Central America and Asia. Half of the Fortune 100 companies have used LLamasoft’s platform for their supply chain designs.

Sid spoke about how to quantify and improve supply chain benefits through simulation and optimization vis-à-vis the use of UAVs for cargo delivery. He cautioned, however, that it still remains to be seen whether UAVs can indeed add real and ongoing operational benefits to existing supply chains. LLamasoft’s goals at this early stage are to quantify the expected benefit and to identify configurations and designs to increase benefits through modelling, optimization and simulation. More specifically, LLamasoft quantifies the cost of transport, infrastructure, inventory holding and expiry while also quantifying performance in terms of response time and stock availability. To this end, Sid explained that LLamasoft aims to answer the following key questions: Which types of UAV platforms should be used? Where should these be located? How frequently should they deliver, to which customers and in what sequence? Which system design variables are acting as binding constraints? He drew on two case studies (Nigeria and Tanzania) to highlight how these questions are applied to supply chain challenges.

Keller Rinaudo (Presentation Slides) from Zipline introduced his company’s pioneering work in Rwanda. Their drone, called a Zip, provides rapid, on-demand aerial deliveries. The Zip has a range of 150 km and a payload capacity of 1.5 kg. Zips drop their payload from an altitude of 5 metres, and then return home.

Each Zip will soon make up to 15 deliveries a day of blood to rural transfusing clinics across Rwanda. Ultimately, Zipline seeks to have Rwanda’s 11 million citizens within a 30-minute delivery of any medical product. Zips do not need a runway to take off or land. They use a catapult and a special method to land within the size of two parking spaces. Zips are operated from a hub, which is a modified shipping container. These hubs are located next to existing government medical warehouses. Each hub, which is operated by 2-3 employees, has a service radius of 75 km. Zipline has secured a strategic, multimillion dollar partnership with a company that allows the company to deploy a hub anywhere in the world within 24 hours in order to support rapid response efforts following major disease outbreaks such as Ebola. Zipline has also struck important partnerships with United Parcel Service and the Global Alliance for Vaccines and Immunization. In addition, Zipline has secured a considerable amount of venture capital funding, making them the most well-funded cargo drone company at present.

Zip

Zipline’s delivery RPAS Credit: Zipline

The RPAS are called "Zips", which can operate from modified shipping containers. Credit: Zipline

The RPAS are called “Zips”, which can operate from modified shipping containers. Credit: Zipline

Dr. Patrick Meier (Presentation Slides) from WeRobotics presented the organization’s mission and Flying Labs model using Kathmandu Flying Labs in Nepal as an illustration. WeRobotics is an independent not-for-profit organization launched by the Humanitarian UAV Network (UAViators) and Drone Adventures. Patrick and his team were invited by Kathmandu University (KU) to provide training and robotics technology (drones) to support humanitarian efforts in Nepal. WeRobotics offered professional hands-on training to KU and other stakeholders on how to use drones safely, responsibly and effectively for disaster risk reduction and early recovery efforts. This included training on best practices, lessons learned, standard operating procedures and codes of conduct. Thanks to partnerships with DJI and Pix4D, WeRobotics was also able to donate 10 DJI Phantoms and free licenses of the Pix4D software to KU and partners. DJI and Pix4D also provided hands-on training on their technology and software solutions. This combined training approach was followed by a joint project with KU and the Community Disaster Management Committee (CDMC) to survey the town of Panga, which had been heavily damaged following the 2015 earthquakes. The CDMC was specifically looking for more detailed and up-to-date data to inform their recovery and reconstruction efforts in Panga.

Together with WeRobotics, KU, DJI and Pix4D, the CDMC carried out aerial surveys using the Phantoms over the course of two days. The resulting very-high-resolution map was then printed out on large banners and brought back to the local community of Panga for a community mapping exercise. The CDMC invited members of the community to annotate the map with their local knowledge, a process that sparked community-wide conversations on the priorities and strategies of the recovery and reconstruction efforts, all informed by the same dataset. Patrick notes that their joint survey of Panga was incomplete, however. But because WeRobotics and the team had taken the time to train their local partners and transfer the technology to them, KU and the CDMC returned to Panga and completed the survey entirely by themselves. What’s more, they carried out two additional surveys, and also created 3D models of a nearby temple and hospital. In addition, KU has been using the Phantoms for an environmental monitoring project they are doing with the World Wildlife Fund.

After sharing the Nepal case study, Patrick gave a more formal overview of the Flying Labs model, noting that the model includes the incubation of local businesses that offer robotics as a service. He announced that WeRobotics had recently secured an important start-up grant from The Rockefeller Foundation to co-create their first three Flying Labs, one each in Asia, Africa and Latin America. He shared his organization’s cargo delivery plans for Nepal, the Philippines and Peru. In closing, Patrick emphasized that WeRobotics is not limited to using aerial robotics (drones) but also actively explores the maritime and terrestrial robotics space.

Alexis Roseillier (Presentation Slides) from Wings for Aid shared their project plans for long-distance, heavy cargo drones. Wings for Aid comprises a coalition of organizations including leading universities and companies in the Netherlands. Wings for Aid seeks to manufacture heavy, gas-powered drones that can cover a range of 500 km with a payload capacity of 100 kg. Their drones will operate along predefined paths, and drop cargo using parachutes from an altitude 15 metres at up to five specified locations authorized by ground staff.

David Guerin (Presentation Slides) from the International Federation of Air Traffic Controllers’ Associations and DLR, Germany’s national aeronautics and space research center, detailed a DLR project examining the potential of low-altitude unmanned cargo aircraft in delivering 1 000 kg to a location 300 miles from the departure point. This project has explored various types of unmanned aircraft in context to complexity of such an operation. He also noted a shift in the regulatory direction provided by the European Commission, and that the European Aviation Safety Agency will be imposing changes throughout Europe. These changes focus on presenting a framework that is operation-centric, proportionate, risk-based and performance-based while being introduced in a smooth and progressive manner, and appear to be the world’s best practices as other countries and regulatory bodies begin to replicate this technique for categorizing operations.

Moving Forward

At the end of the meeting, participants were asked to reflect on next steps that they would like to see implemented. One suggestion proposed periodic follow-up meetings. This proposal prompted the idea that any potential follow-up discussions would benefit from more field deployment data to analyse the impact of field developments. It was further suggested that the humanitarian community stay involved in future discussions concerning the application and evolution of cargo drone uses in the development community.

Annex I: Agenda

Day 1 – June 20, 2016

14:00 – 14:25: Welcome and Keynote Address
Professor Lenny Koh (University of Sheffield) PDF

14:25 – 14:35: Introduction to Drones in Humanitarian Action | Framing and Outputs
Denise Soesilo (FSD) PDF
Romain Gauduchon (DG ECHO)

14:35 – 14:45: Cargo UAVs as Humanitarian Robotics
Dr. Patrick Meier (WeRobotics) PDF

14:45 – 15:15: Presentation of White Paper: UAV Humanitarian Logistics
Daniel Ronen (UAVaid) — Link to the whitepaper is forthcoming

15:30 – 16:10: Speaking Slots 1&2 + Open Discussion: Humanitarian Needs (2 x 10 min. talks)
Joel Kaiser (Medair)* PDF
Rae McGrath (MercyCorps) PDF
16:25 – 17:05: Speaking Slots 3&4 + Open Discussion: Logistical/CBA Needs (2 x 10 min. talks)
Jonathan Ledgard (Redline)* PDF
Luca Palermo (World Food Programme – Air Transport Unit) PDF
17:20 – 18:00: Speaking Slot 5&6 + Open Discussion: Technology Needs (2 x 10 min. talks)
Oriol Lopez (MSF) PDF
Florian Schneider (Quantum Systems) PDF
19:00 Conference Dinner

Day 2 – June 21, 2016

08:45 – 09:00: Welcome & Overview of Day 2
09:00 – 09:10: Welcome by Professor John Derrick (University of Sheffield)
09:15 – 10:15: Lightning Talks
Sid Rupani (LLamasoft)* PDF
Keller Rinaudo (Zipline)* PDF
Patrick Meier (WeRobotics) PDF
Alexis Roseillier (Wings for Aid) PDF
David Guerin (Int’l Federation of Air Traffic Controllers’ Associations) PDF

11:00 – 12:00 Breakout Sessions

Session 1: Regulations
Facilitator: David Guerin (Int’l Federation of Air Traffic Controllers’ Associations)
Session 2: Big Data & Business Models
Facilitator: Lenny Koh (University of Sheffield)
Session 3: Humanitarian Technology Needs
Facilitator: Oriol Lopez (MSF)

12:00 – 12:30: Report Back
12:30 – 12:45: Pick up Lunch
13:00 – 13:45 Breakout Sessions
Session 1: Code of Conduct
Facilitator: Patrick Meier (WeRobotics)
Session 2: Realizing Supply Chain Improvements
Facilitator: Ian Shellard (University of Sheffield and UAVaid)
13:45 – 14:15: Report Back
14:15 – 14:30: Wrap-Up & Next Steps

* Presentations delivered via Skype

Annex II: List of Participants

 

AerOptim Fabien Bonnassie
Airbus Helicopters Tim Williams
DG ECHO Romain Gauduchon
Emergency Telecommunications Cluster WFP Haidar Baqir
FSD Denise Soesilo
FSD Valeria Fabbroni
Handicap International Paul Vermeulen
Imperial College London Jose Escribano Macias
International Federation of Air Traffic Controllers’ Association David Guerin
Map Action Mark Gillick
Medair Joel Kaiser
MSF Oriol Lopez
PRIO Kristin Sandvik
Quantum Systems Florian Schneider
Redline Jonathan Ledgard
TH Pharmacy Trudi Hilton
UAVAid Daniel Ronen
UAVAid Ian Shellard
University of Manchester Joseph Lampel
University of Sheffield Lenny Koh
WeRobotics Patrick Meier
WFP Jose Odini
WFP Luca Palermo
World Bank Edward Anderson

 

Annex III: Discussion Group Notes

Code of Conduct

Facilitated by Patrick Meier (WeRobotics)

Patrick started the session by introducing participants to the existing International Humanitarian UAV Code of Conduct, which he spearheaded with the Humanitarian UAV Network (UAViators) in close collaboration with dozens of humanitarian organizations. The Code of Conduct was first launched in early 2014 and revised extensively over the course of two years through an open multi-stakeholder approach supported by the Rockefeller Foundation. The Code of Conduct also includes Supporting Guidelines on priority topics including Data Management, Community Engagement, Effective Partnerships and Conflict Sensitivity. Patrick noted the need to include Cargo Delivery to this set of guidelines, hence the purpose of the session on Code of Conduct.

Patrick invited participants to suggest guidelines to inform the safe, responsible and effective use of cargo drones in humanitarian settings. Participants made recommendations based on actual and perceived risks in using cargo drones, including, for example, managing risks related to dropping heavy loads. In addition, the contents of the cargo, which must be based on actual or expected needs identified by formal humanitarian organizations, can pose risks that need to be managed accordingly. Cargo delivered by drones – like with any other mode of transportation – must follow humanitarian principles and thus not discriminate or marginalize at-risk communities. Organizations must also have good reason to use drones over other, potentially more reliable and effective transportation mechanisms.

The transportation of cargo by drones must follow relevant national and international standards and laws governing the transportation of sensitive or dangerous materials, including medication and vaccines. If a cargo drone uses a parachute or airbag to cushion the drop of cargo, the parachute or airbag must not contribute to waste or environmental risks. Ideally, these should be recyclable or easily used locally to meet other needs. The discussion group also considered risks related to accidents, including environmental risks and noise pollution. That accidents or technical problems can prevent urgent payloads from being delivered means that contingency plans, fail-safe mechanisms and redundant systems are necessary. Companies and organizations operating cargo drones must have drone recovery procedures in place, and publicly report any incidents caused by or related to their drones, which should be equipped with the equivalent of a black box. Operators of cargo drones must also have relevant insurance.

Participants also discussed risks associated with route planning and flight paths, which are typically in fixed corridors approved by aviation authorities. Armed groups could therefore learn of these routes and intercept the drones and commandeer the cargo. In any event, procedures must be in place to quickly reroute flights as needed regardless of whether the hazard is man-made or not. Participants also encouraged the use of open standards and open data practices to share relevant data publicly when doing so does not violate standard data protection protocols. Local communities whose immediate airspace is used to transport goods by drones ought to have access to data relevant to these air corridors. In addition, any plans to establish air corridors should take into account any harm that could fall on local communities living under these corridors. Ideally, local community representatives should be consulted and informed about these plans and have access to a complaint mechanism should any concerns arise after the corridors become operational.
Finally, participants discussed the possible rise of cargo drone monopolies versus the operational complexities of enabling multiple companies (with different types of drones) to operate in the same complex airspace. The main concern is that the first company that wins government approval to set up services in a given country may ultimately enjoy a monopoly because adding other companies and their drones to the same air corridors will be a complex technical challenge. Should the first company be obliged to share air corridors? Should air corridors be like the audiovisual spectrum? Should they be auctioned? Should part of the “spectrum” be provided free of charge to humanitarian organizations? Patrick suggested that the complexity of managing different drones manufactured by different companies could be simplified by simply segregating air corridors by time. In other words, you can only have one drone in any given air corridor at a time. Only when that drone has landed can another drone operate in that air corridor. Ultimately, a system may be needed to enable dozens of different drones to operate in any given air corridor, but there needs to be obvious demand for such a high volume and frequency of air cargo delivery.

Humanitarian Technological Needs

Facilitated by Oriol Lopez (MSF France)

This session analyzed the technology needs of humanitarian organizations, discussed regulatory options and considered potential technological solutions given the current and prospective needs and limitations in the field. Participants noted that cargo drones would need to compete in the existing supply chain, which includes:

  • Motorbike delivery
  • Road Infrastructure
  • Airports and manned craft air delivery
  • Other

The session identified the following regulatory constraints for cargo UAVs by categories:

  • < 25 kg aircraft -> 2-5 kg payload, 40-120 km
    • Compete with: Motorbike delivery
    • Operated mostly by end users (NGOs)
      • Few operations per week up to 4/day
      • Low requirements of operator
      • Highly automated
      • Self safety features
    • Operated by small packet delivery companies
      • Several operations per day
      • Higher operator requirements
    • 25 kg aircraft -> 20-100 kg payload, 200-800 km
      • Compete with Land Cruiser, small trucks, small helicopter, small plane
        • High rotation, start with 20 kg increments
      • Operated by WFP or similar
        • Higher pilot and infrastructure requirements
      • Operated by NGOs only in specific cases (high need, low volume)

Examples of potential uses brought up during the discussion include the following:

  • Drop < 5 kg during first stage of emergencies – natural disasters
    • Delivery of cell phones, satellite phones or other high-value items
    • Range and payload: Few kg (2-5 kg) to few km – < 80 km range
    • Aircraft:
  • Fixed-wing craft with Vertical Take-Off and Landing (VTOL)
  • Long-range copters (min 20 km)
  • Fixed wing + parachute drop
  • Pickup < 5 kg lab diagnostics and research samples – stable contexts
    • VTOL
    • < 200 km range
  • Drop > 20kg – Emergency Response Non-Food Items (NFI)
    • Blankets, water, tents, medical equipment
    • <100 km
    • Need to operate without airstrips (VTOL) or very short airstrips
  • Refugee camp drops – one per day, no airstrip or short (200 m) airstrip
  • Aircraft: Fixed-wing or VTOL
  • Cargo: 25-200 kg
  • Range: 150 km
  • Mass vaccination campaign
    • Cold chain
    • Delivery on demand to the teams depending on need
  • Air ambulance (concept)
    • Carry patient
    • Payload: 120 kg
  • Range: 100 km

Participants discussed the need to develop a curve of cost and performance for electric versus petrol aircraft, and developed the following criteria to be considered:

  • Electric
    • Limit VTOL – 50 kg take-off
  • Limit electric plane – unknown
  • Fuel Cell (3X battery density), very expensive (Quantum Systems looking into it)
  • Hybrid
    • VTOL
    • Electric vertical, petrol horizontal?
    • Alternator plus all-electric motors?
    • Alternator: No batteries required, fewer issues at customs

 

Regulations
Facilitated by David Guerin (DLR)

With 19 annexes and over 10 000 standards and recommended practices (SARPs), the Chicago Convention on International Civil Aviation is extremely complex. The previous five years have been key for UAV-specific policy development, and the next two years will bring additional major developments. The regulating body responsible for international aviation, the International Civil Aviation Organization (ICAO), is tasked with ensuring safe and efficient aviation. ICAO does not yet stipulate regulations for autonomous or low-level operations but only for cross border operations. Thus, member states are currently formulating their own regulations, an approach that leads to a patchwork of different policies and a lack of standardization between countries. Amending SARPs can take five to seven years, while global implementation of new rules can take decades, and differences will still exist in several countries.

Recently, the European Aviation Safety Agency (EASA), under the direction of the European Commission, has expanded its regulatory role beyond its previous mandate of operations heavier than 150 kg across Europe. EASA is becoming responsible for all unmanned operations in Europe and is currently taking a lead to adapt regulatory frameworks to the rapid adjustments that need to be made to accommodate Remotely Piloted Aircraft Systems (RPAS or UAVs) in a way that is constructive and safe. It is focusing on six guiding principles:

  1. Operation-centric: In the past, regulations for conventionally piloted aviation have focused on whether the aircraft will work or fall out of the sky, on the relationship with other airspace users, and on certification and airworthiness. Looking at the actual operation is a big change in focus.
  2. Risk based: The ICAO Safety Management Manual (Document 9859) provides guidance in the establishment and application of an operator’s safety management system. The European approach, however, is risk-based, where the risk to people on the ground or damage to infrastructure is examined separately from the airborne risk to other airspace users. This approach considers areas such as the type of airspace, the number of aircraft in and area and population density. Also, all operations are considered regardless of who operates the aircraft. Some member states have considered commercial or individual actors separately in the past.
  3. Performance-based: Looks at how systems will perform, and moves from methods or technical solutions to outcomes.
  4. Proportionality: By establishing three categories of operation (Open, Specific or Certified) the complexity of the operation is classified, for example, removing some restrictions on low-complexity and low-risk operations while matching conventional aviation regulations for big organizations with aircraft operating like piloted aircraft or in a complex or high-risk manner.
  5. Progressive: Allowing operators to step up from the lower to the higher categories.
  6. Smooth change: A seamless transformation that does not introduce an undue burden for other aviation stakeholders.

Eurocontrol, the agency tasked with aligning air navigation services in Europe, is affiliated with EASA and the EC in progressive RPAS regulations. The agency is avoiding prescriptive regulations while aiming to have buy-in from all European stakeholders through a functional concept of operations for RPAS that emulates the EASA direction. Most likely, European countries will adopt the EASA and Eurocontrol recommendations, and other countries and regions may embrace the European policies either individually or if ICAO introduces supporting SARPs. The EU appears to have taken the global lead although it is not the only centre of excellence in progressive regulation.

The UAV industry is changing at dramatic rates and can be seen to reinvent itself every eight years. This is vastly different from conventional aviation. A secondary distinction is that many participants in RPAS operations are not typically from an aviation background. An important focus is the determination of the use of the lower altitude airspace. A few large companies with great resources and social support through social media and expertise may try to influence regulations to their advantage. The contemporary aviation players, the airlines, aircraft manufacturers, air navigation service providers and regulators appear to be waiting for direction from bodies with technical expertise while the latter look for guidance from the former.

An historically conservative approach may not stand up against the enormous number and diversity of RPAS, both operations and platforms, and the weight of the bodies demanding change. The present regulatory change framework needs to be revamped to achieve a more proactive methodology. Countries are considering the advantages of the introduction of three levels of rules – hard regulations at the higher levels, covering basic principles and essential requirements; regulations addressing implementation of the rules; and a third level providing guidance material, specifications and allowable techniques for complying with the rule set.

Dropping from drones: Being able to drop cargo is potentially a very important feature in cargo deliveries, but the dropping of items is not currently allowed in many countries and may not be considered when regulations are revised unless the value to humanitarian aid groups is articulated.

Safety: Safe operations are key to future integration of RPAS, and individual countries are often left on their own to assess and approve an operation without internationally harmonized rules. The consequences of a serious incident or accident involving RPAS and other airspace users or the public are probably dire, and this prospect necessitates a safety focus for all operations to avoid unsafe undertakings and to manage risks. Operators need to introduce appropriate approved safety systems such as drones that are frangible and visible (to the eye and to other aviation surveillance) and have geo-fencing systems to control malfunctions within an agreed area of operation.

BVLOS: A line of sight requirement is a major challenge for cargo delivery. Many countries do not allow or severely restrict operations of drones Beyond Visual Line of Sight (BVLOS). Visual Line Of Sight (VLOS) limits are very close, and most operators want to go beyond these limits, especially if the aim is to deliver cargo over long distances.

Sense and avoid/detect and avoid technology (SAA/DAA): Sense and avoid technology is a method of conflict avoidance in airspace and is expected to replace the pilot’s visual vantage point from the cockpit. Presently, this technology is complicated and expensive, takes up a lot of power and existing “reliable” systems are only on large, complex and expensive technology. Some SAA/DAA technology could be adopted from conventional aircraft systems although a new approach is more likely to succeed. Equipping all RPAS with SAA/DAA will be unachievable. The safety issue is not restricted to avoiding other aircraft but also in avoiding flight through clouds and turbulence, and terrain and obstacles on the ground.

Insurance and liability of RPAS operations are very important sidelines to the regulatory discussion.

Big Data and Business Models

Facilitated by Lenny Koh (University of Sheffield)

Expertise

  • Healthcare
  • Logistic and supply chain
  • Humanitarian
  • Innovation
  • Big data
  • Mining
  • Drone
  • SOP

Stakeholders

  • Global Humanitarian Lab
  • European Space Agency

Business model

  • Logistics industry set up permanent team
  • Common service
  • Drone as a service
  • Preparedness before, during and after disaster

Resources
For an overview of the Initiative on Drones in Humanitarian Action: https://drones.fsd.ch/
Sheffield University: http://www.sheffield.ac.uk/
Meeting documents and presentations:https://drones.fsd.ch/2016/06/23/cargodelivery-drones-in-humanitarian-action-a-first-meeting-summary/

[1] Grant title: “Fostering the appropriate Use of Air Borne Systems in Humanitarian Crisis”
[2] Case study: https://drones.fsd.ch/2016/04/13/using-drones-for-medical-payload-delivery-in-papua-new-guinea-case-study/

Report prepared by Patrick Meier, Denise Soesilo, David Guerin
Copy edit: Geoff Hughes

Drones in Humanitarian Action is partially funded by DG ECHO. This website covers humanitarian aid activities implemented with the financial assistance of the European Union. The views expressed herein should not be taken, in any way, to reflect the official opinion of the European Union, and the European Commission is not responsible for any use that may be made of the information it contains.