Should Humanitarians Operate Satellites for Their Missions?
While at SatSummit this past May, a group of us debated if “the humanitarian sector should operate its own satellites.” Rhiannan Price, and myself (Bill Greer), argued in the affirmative, while Dr. Mark Iliffe and Dr. Ziad Al Achkar argued in opposition. Shout out to Akesh Mallia for moderating ‘The Great Debate’. After an hour of discussion, with both sides making strong arguments for their position, the question remained unresolved. I wanted to take a moment to walk through the arguments, for and against, in hopes of starting a greater conversation on the topic, in addition to sharing the highlights of our discussion with a wider audience.
Arguments for dedicated satellites in humanitarian aid and disaster response
Earth Observation data has a long history within government agencies around the world, specifically involving its use in intelligence, surveillance, and reconnaissance (ISR) missions. This data is critical to the success of these and many other government missions. While humanitarians have both different mission requirements and desired mission impacts, the potential value to the community has been proven over and over again in project-based successes. Satellite imagery has been used to:
- Monitor atrocities such as the Rohingya Massacre in Myanmar and Sudan
- Fight human trafficking and Forced Labor
- Monitor the growth of Refugee and IDP Camps
- Show the damages of war in Iraq, Ukraine, Gaza, and Others.
- Show the damages from disasters such as wildfires, flooding, earthquakes, and others.
These are just a few examples of commercial satellite companies using EO data for humanitarian aid and disaster response. Earth Observation data could play a greater role in solving these problems at scale, above and beyond the individual successes just shared. The commercial satellite industry, however, is primarily focused on defense and intelligence use cases, followed secondarily by commercial use cases, such as precision agriculture and property insurance. While the data has proven impactful, humanitarian use cases tend to be an afterthought for commercial companies. Tasking high-resolution satellites for humanitarian cases is nearly always in conflict with defense customers’ needs to monitor active conflict zones or other commercial interests. This results in a lack of consistent and dedicated access to satellite imagery for humanitarian crises. Further, understanding the licensing behind that data is critical to having success in the field. If your organization can’t predict access to data, how can you use it to effectively accomplish your mission?
Several commercial companies support humanitarian aid and disaster response through “open data programs”, such as Maxar, Planet, Satellogic, Umbra, and others. These organizations and others are also supporting data-sharing efforts through their work with the International Disaster Charter. These “Open Data programs” are, however, 1) still dependent on the goodwill of commercial companies, 2) limited in scope, i.e. typically just to disaster response, and 3) do not touch on planning, mitigation, and recovery activities. This limits humanitarian practitioners from being more proactive and creating a system for consistent improvement. Open data is also often at the capricious discretion of these companies. Which disasters make the cut to have open data released? When and how much data is released? These commercial companies have exquisite (or powerful?) capabilities, but their licensing models often leave them in conflict with supporting humanitarian aid and the distribution of imagery. We have also seen increased pressure for commercial companies to show profitability, putting these less profitable missions at risk, as well as threatening a loss of these invaluable resources should any of these companies fail.
A dedicated resource is needed for success
A purpose-built satellite constellation focused entirely on humanitarian use cases would provide the community with more consistent access throughout the lifecycle of humanitarian and disaster events. This would critically extend the ability to cover more events, including those that do not receive much attention in international media. This would allow for skill and capacity building by the community to consistently improve over time, at scale, which is discussed further below. A humanitarian satellite would pave the way as a global tool to be used across the sector where it is most needed.
Disasters are increasing in both occurrence and scale of impact on human populations. Natural disasters on a global scale have increased 10x in the last 40 years, and ‘billion-dollar disasters’ are also seeing a sharp increase, with 28 total disasters each with over $1 billion in damages in the US alone (2022). The nature of this challenge is that it cannot be solved locally, because the scale of the problem is global. The International Disaster Charter had 49 disaster activations in 2021, 51 in 2022, and 63 in 2023. This is just another example of the extent of the continuously growing challenge. Current commercial EO data business models do not effectively address this need for this community.
The Landsat and Sentinel programs have proven there is a huge demand for data, and when it is made open, it is used and adopted to great effect. The data has proven itself to be one of the key sources of data for understanding climate change and changes to landmasses. The downside of these global monitoring missions is that the resolution is not good enough for viewing the types of human activity that are of most direct interest to humanitarians. Meanwhile, commercial companies operate high-resolution, taskable satellites, but tasking a satellite is subject to both market pressures and profit margins making it difficult for humanitarian organizations to outbid government defense and intelligence organizations for imagery when needed. There is a need for higher resolution imagery to support the growing demand, but no dedicated resources for this mission, resulting in both a market and a public good gap.
The status quo does not improve humanitarian outcomes
Can’t we just keep tasking commercial satellites? The commercial market provides many solutions, from super high-resolution imagery from Maxar’s newly launched Legion satellites, Planet’s Pelicans, Satellogic, BlackSky, synthetic aperture radar (SAR) from Umbra, Capella, and IceEye, multi and hyperspectral data coming from Wyvern and Pixxel, and thermal data from Sat Vu, among countless other missions both current and future. We have an extremely robust set of capabilities in the commercial market that are looking to gain a foothold to become profitable. These companies all share a common business model of tasking imagery for the highest bidder — this is how the satellite imagery business works. The defense and intelligence market is nearly always the highest bidder, but when they do not have an interest in an area, collection priority moves on to the next interested party. These satellites are visionary tools and resources that are shared across multiple markets, but they are not dedicated to a specific mission or use case. The conflicts of interest that exist when you task for cash cannot be overcome through any for-profit satellite. The mission always devolves to the winner of the auction.
We are seeing new players emerge on the mission-focused satellite side of the industry, creating new opportunities in the space, and working towards an aspirational vision of the future. Some exemplary examples are MethaneSat and Firesat. These companies are proving that public and private entities can work together towards the shared mission of improving the world and solving the problems of methane emissions and wildfire detection, respectively. This shows that focused use cases for satellite constellations are gaining traction, and the industry is starting to focus on solving specific, narrowly scoped problems while moving away from the Geospatial Product Trap (as coined by Will Cadell). These satellites are focused on changing current business models to achieve their goals, and not just when there is a big news story. These organizations will create a more holistic understanding of emissions and wildfire throughout the world for academics, commercial companies, and public benefit alike. They are not tasking for profit, like the rest of the industry, but they are tasking for a sustained mission over time.
Decreasing prices and technical barriers make now the time to execute
Satellite build costs and launch costs are coming down, resulting in a significant spike in the number of satellites in operation as a result. Over the coming decade, it is estimated by Euroconsult that there could be 15,000 satellites in orbit, including a 3x increase in earth observation satellites. None of these satellites will be solely focused on humanitarian use cases, with most dedicated to defense and intelligence, emissions monitoring, agriculture, or new technology demos. Further, the cost of building and operating satellites has decreased over time, while the need for the data they produce continues to grow. Meanwhile, the cost of tasking imagery is proving to be fairly inelastic due to these commercial companies needing to make fairly high margins on their highly capital-intensive startup costs. While imagery may be discounted drastically when not in conflict with other buyers, the entire process is very opaque, making it difficult to understand the frequency and impact of such discounts. Increased competition among bus manufacturers and payload providers gives commercial companies more options, in turn creating a more competitive market.
Working with Earth Observation data is also becoming easier (for those who are already using it). AI, cloud-native tools, and greater standardization in data processing make it easier to approach and use huge amounts of satellite data with remarkable outcomes. Companies such as Felt Maps, Clay, Fused.io, and Seer.ai can access, mash up, display, and distribute insights where needed, without massive technical teams. ESRI has the largest footprint in the industry and is improving its raster data storage and processing systems to enable Earth Observation workflows at scale while delivering to the masses. The last mile is still the most complex part, but there is a whole ecosystem making drastic improvements to the status quo. These improvements can make it much easier to work with big data. Users with consistent access to big data are fusing diverse data sets in amazing ways, but without access to consistent data and the ability to implement these tools humanitarians will fall further behind in the widening digital divide.
Adoption of satellite imagery in the humanitarian space requires openness and consistency
Humanitarian organizations gain a huge amount of public support from the general populace regarding their aiding in disasters, but that attention is as fleeting as the news cycle. Similarly, the open data commercial companies provide is short-lived and often burdened by difficult licensing terms. As a result, much of the open disaster data that exists today many organizations do not utilize because it is too ephemeral, and use restrictions make it unviable. Tools such as OpenStreetMap and the Humanitarian OpenStreetMap Team (HOT) have shown value in crowdsourcing, bringing new users in to help when licensing allows. HOT has engaged with nearly 350K users to help map out buildings and roadways in areas of humanitarian need. Recently, both of these vibrant communities lost access to Maxar imagery as a baseline data set for their work. The dependence on the unpredictable charity of commercial companies drastically cripples these projects’ ability to execute their missions in disaster situations and long-term humanitarian crises.
The Landsat and Sentinel missions also saw massive increases in utilization of their data when they moved from a paid model to an open data model. This shift in utilization is well covered by Joe Morrison in his provocative blog post. Landsat’s data caused a deluge of innovative research never seen under the “pay to play” model. “As of February 2022, there were 767,000 publication records in Google Scholar listed with the keyword Landsat.” (Wulder, et al, 2022). Not only did the free and open data policy expand data utilization, but it also increased the depth and scope of the questions asked. Open access to data enables researchers to address new questions and work across national and disciplinary boundaries.
While Joe Morrison calls for opening of archive data, I believe a donor funded humanitarian mission could both open the archive, as well as tasking, creating a new model for the industry. A humanitarian-focused satellite mission, built to be open in its data sharing, and tasking, would be a force-multiplying tool for current humanitarian practitioners. This mission would also introduce a larger community of academics, students, commercial users, and donors to the value of the data, which would be solely focused on areas of humanitarian need.
Arguments against a dedicated satellite for humanitarian aid and disaster response
Financial aid could be better used elsewhere
Costs are indeed coming down. Ten years ago the cost of a similar capability in space was in the $100M price range, and it is now in the $10M range. Further, we now have many more options on how to create that platform. We must still look to see if this is the best possible use of that money. With most humanitarian crises chronically underfunded, $10M could be spent to help feed the hungry, grow crops, and improve living conditions in refugee or IDP camps. The return on investment could be linked more directly to the impacts.
The ability to execute with new technologies is limited
Using satellite imagery for humanitarian aid and disaster response has yielded impressive results in the past, but people don’t see the huge amount of experience and effort that it takes to get to those results. GIS, spatial analytics, and getting important and relevant answers from satellite imagery is no simple task. It requires specialized processing and knowledge of the subject matter and location you’re trying to analyze or understand. The humanitarian community has very few individuals with the skills and expertise needed to effectively use a dedicated satellite constellation at scale.
Satellite operations are not something that humanitarian groups should also be responsible for
Building and operating satellites may be getting easier, but who is responsible for the satellites, and what business model would they follow? Humanitarian organizations are already accountable for such a wide range of activities, asking them to take on the operations of a satellite constellation doesn’t play to their strengths and divides focus. Large Humanitarian organizations are not well equipped to run tech startups, especially not hardware in space. The expertise does not exist within the community to own, operate, and execute a satellite mission. The Environmental Defense Fund (EDF) was able to execute MethaneSat and the Earth Fire Alliance is making progress on FireSat, however, those problems and solutions do not directly transfer over to Humanitarian organizations, and new organizations were developed to manage the operations of these satellites.
Outside of the general difficulty of building and operating a satellite, or constellation, there are also complications in what you task the satellite to image. For an open satellite model to work, the community of interest would need to develop a new governance structure to decide what to task and to make tradeoffs that are not based on which image will be most profitable, but rather, what is most valuable. Defining the scope of the mission, and what AOIs are most relevant to that scope can be ambiguous at best. Distribution of the imagery is also less than straightforward, both technically, and from a licensing standpoint. Ensuring distribution will not have negative externalities while taking advantage of the potential upside of that imagery is a nuanced and difficult job. There are also political, ethical, and safety concerns about opening up data to the public or restricting access, which inherently becomes political. Ethical and safety concerns can vary widely by region and should have local input into how the data is used and where it will have the greatest impact. These concerns are further exacerbated when we look at the use of satellite imagery in ongoing conflicts where humanitarians are actively operating. Potential perceptions of humanitarians’ use of assets that are typically defense and intelligence assets could have negative unintended consequences, depending on the environment in which they are working. This host of problems would all be added on top of the already full plate that most humanitarian organizations need to deal with.
Current commercial partnerships exist to get data to humanitarians
Earth Observation data already exists within the commercial market, and the community can acquire that data just like any other customer. As mentioned above, commercial operators provide open data for large disasters for free, and users can still task the same satellites to fill any data gaps. While tasking is costly, it is a low-risk way to ramp up humanitarian capacity. The key to improving the adoption of satellite imagery by humanitarian organizations lies more in fixing the issues with licensing than creating new satellites or assets. Organizations with imagery needs should be able to test hypotheses and measure results based on imagery currently available through commercial open data programs, or through buying imagery from the commercial market.
Previous attempts
There have been previous attempts for humanitarians to use satellite imagery for specific events. In 2010 the Satellite Sentinel Project (SSP), spearheaded by TheSentry.org, the Harvard Humanitarian Initiative, DigitalGlobe (Now Maxar), and a group of high-profile celebrities, such as George Clooney, used satellites to focus on war crimes and atrocities in Sudan. The project had limited impact due to the inability to apply consequences to perpetrators of war crimes. These limitations tie into the last mile problem for satellite imagery and its effectiveness in the field. I should also mention that there were many successes from this mission, “These successes resulted from unique insights that only remote-sensing technologies could deliver; improving the speed, agility and tempo with which evidence could be collected and presented for the benefit of vulnerable populations.” (While We Watched: Assessing the Impact of the Satellite Sentinel Project.) The SSP project was undertaken using tasking from commercial companies, which has been well addressed above. This was however hugely successful in calling media attention to the crisis in Sudan.
Points of Agreement & Next Steps
Despite these challenges, there is a consensus that the humanitarian sector needs improved access to satellite imagery and data. Owning and operating a satellite is well within the realm of possibility, however, for the satellite to have the impact to be viable, it would take significant changes to the current business models. The data is valuable and has the potential to make a difference on the ground. Collaborations between tech companies and humanitarian organizations can pave the way for shared resources, reducing costs and improving operational capabilities. Partnerships could focus on creating a framework for data sharing, training, and joint missions.
GIS and remote sensing skill gaps still exist even with the improvements in systems and tools. Global development and Humanitarian organizations employ amazing people from all over the world, with unique skills that can improve outcomes locally, but global-scale tools and monitoring are needed for some of the largest problems. Cross-functional growth is required to fill these gaps, which can be accelerated through access to the proper data, tools, and systems.
Whether humanitarian organizations should operate satellites is complex, balancing the promise of enhanced response capabilities against the realities of cost and expertise. There is potential to greatly improve life for the vulnerable with these technologies. There are challenging issues to address across the board. The satellite imagery business needs to address licensing, accessibility, and expertise, while the humanitarian sector needs to address the politics, safety, and responsible spending of humanitarian funds. Building, launching, and operating a satellite for this community is not the finish line, but the start to improving systems for humanitarian organizations.
If we as a community want to make progress toward the development of a dedicated satellite system we must:
- Define mission with users on the ground: satellite imagery is only as good as the people using it, and narrowly focuses on the use cases that cannot be solved using other means.
- Commitment to solving specific problems: A humanitarian satellite initiative would need a coalition of organizations committed to supporting and solving problems in which they can provide expertise on the ground. The lack of a distinct mission and use cases is an important component of making a useful tool for the greater community. Everything from the operational model, data collected, licensing, and data distribution needs to be in service to a specific set of missions.
- Define metrics of success for a humanitarian satellite mission: defining the metrics where a humanitarian satellite would have an impact and where it would not are important details to determine the value of the mission. Clarity on metrics and goals must be shared by the coalition of supporters.
- Accessibility of the imagery: making sure that the imagery is available to the broadest possible user base and tasking is available to the areas of greatest need to execute the mission objectives is key. A humanitarian satellite mission needs to be used as a tool for the entire community. It should be a shared resource / public good that allows for the proliferation of the knowledge it collects to be distributed to the areas of need.
- Understand legal, ethical, and political implications: the legal, ethical, and political implications of a taskable satellite for humanitarian missions is a relatively new area. Legally, building and launching satellites is fairly well understood, and groups such as The Locus Charter, have created a framework for the ethical use of geospatial data. The political ramifications, though, are very complex and situational.
If you’re interested in digging deeper into this topic, you’re in luck. SatSummit Lisbon and SatCamp are fostering grassroots innovation and making things happen for the best possible future. These gatherings bring together thinkers and doers from various backgrounds to brainstorm solutions, share best practices, and cross-pollinate ideas and solutions. They serve as incubators for ideas that could ultimately shape the future of humanitarian aid and the satellite industry.
