Development Seed

Subscribe to Development Seed feed
Updated: 2 hours 14 min ago

Bem-vinda Caroline Portugal

Tue, 03/24/2015 - 10:00



Welcome Caroline! Designer and architect Caroline Portugal is joining Development Seed.

Caroline is going to build thoughtful, beautiful software. Caroline hails from Brazil, where she studied and practiced Architecture and Urban Planning. She worked on architecture projects in Brazil and the US ranging from Rio’s Olympic Park Master Plan to the National Museum of African American History & Culture. Caroline’s latest work has been in visual design. She recently completed a graduate degree in Graphic Design from MICA.

Caroline’s work is moving. It is visually stunning. It is grounded in solving practical challenges with humanity, creativity and fun. Check out her work on bike lanes and the future of contraception to get a sense of why we are delighted to have her on the team.

Caroline speaks fluent Portuguese. Say bem-vinda to Caroline on twitter.

Powering Landsat Powertools

Thu, 03/19/2015 - 14:00



Amazon Web Services just opened Landsat on AWS, a publicly available archive of Landsat 8 imagery hosted on their reliable and accessible infrastructure. This investment by the AWS open data team has a big impact on our work to make satellite imagery more accessible.

Our open source Landsat browser, Libra, now has options on some scenes to download individual bands related to specific types of imagery analysis like NDVI, or Urban False Color. The most recent Landsat–8 images are now available for download up to two days sooner. Last week we rolled out a new version of landsat-util, our open source utility for processing Landsat imagery. The new version is much faster and allows you to build false color composites on the fly. These improvements to Libra and landsat-util are possible because we started using Landsat on AWS, which is a publicly available archive of Landsat 8 imagery hosted on Amazon S3 that is publicly available today.

Our newest releases of Libra and landsat-util utilize Landsat on AWS for 2015 imagery. Landsat on AWS provides 2015 imagery as unzipped individual bands. AWS makes this imagery available extremely quickly, often within hours of capture. We can pull only the data that we need and to work with it immediately.

Landsat 8 imagery is an incredibly powerful resource. It is some of the most valuable open data produced by the US Government. Our partners rely on Landsat data for everything from evaluating droughts to tracking conflict. However, until now, individual bands of Landsat imagery has never been available via predictable download endpoints that we can integrate into our tools.

Libra and landsat-util now allow our partners to get imagery sooner and process it faster. Speed and ease are critical to our partners who use this data to respond to natural disasters, prevent hunger, and monitor elections.

Thanks to the AWS team and collaborators–Frank Warmerdam at Planet Labs, Charlie Loyd at Mapbox, and Peter Becker and others from ESRI–for building a Landsat archive with developers in mind.

Libra and landsat-util are open source and on Github. Go ahead and fork or contribute.

Launching v0.5 of landsat-util

Tue, 03/10/2015 - 15:30



We just released a new version of landsat-util, version 0.5. This version is lighter and has fewer dependencies. Landsat-util v0.5 downloads and processes images much faster and gives users more control.

Installing landsat-util now is considerably easier for Mac and Linux users. We are still working to make it as easy to run on Windows.

This new version reflects a significant change in our thinking on landsat-util. We've moved to using faster and simpler frameworks to optimize processing. We removed heavy dependencies, such as ImageMagick and Orfeo Toolbox, that were causing installation problems. We leveraged faster processing frameworks like Rasterio, numpy, and scipy. These changes significantly optimize disk and memory usage resulting in faster and less error-prone processing. Landsat-util can now process images 3x faster.

We'll post more of the technical details on how we rebuilt landsat-util and what's under the hood. In the meantime hit us up at FOSS4G in San Francisco all week to learn more.

Collect and verify mobile reports

Tue, 03/10/2015 - 14:30



A common scenario for mobile reporting looks like this:

  • A group wants to collect reports from their own trusted network and also to crowdsource reports from the public.
  • The group has some process to try to verify crowdsourced reports and needs to track verification.
  • The group wants to publish this data using simple visuals that answer their core question and invites comparison.

We recently worked with a partner looking to do exactly this. With so many tools for mobile data collection, you'd think that this should be easy to do with open source tools. It's not. Here is our experience.

Collecting Data

OpenDataKit is a great open source tool for data collection on Android. When you have control over the device your volunteers are using, ODK is great. (Side note: If you like ODK, check out OpenMapKit an exciting project of the American Red Cross, SpatialDev, and Ona).

When you are relying on reports from volunteers with their own phones you can't always get them to download an Android app. This is why the ODK ecosystem and tools like Enketo and Formhub are interesting. Enketo allows us to serve a simple web form on any device. Formhub gives us the ability to convert and manage Excel-based forms. While this ecosystem is great for authoring forms, we found that this system had some shortcomings:

  • managing multiple datasets can be a challenge,
  • API lacks some features,
  • system is resource hungry, and
  • difficult to deploy and maintain.

We used these tools for what they are best at, generating web forms from Excel documents. We pulled the data via the API to use in other open source tools.


No tool that we looked at had good workflow for verification. Most had no verification or had only a simple yes/no toggle. To make the verification of thousands of reports manageable, the system must have a workflow for quickly reviewing whether reports are verified true, verified false or are unverified.

We used Django to build a verification platform that is useful and usable for the data verifier who is triaging hundreds of reports a day and for the field organizer who is trying to track the status of single report.


To publish and visualize the collected data, we designed a map and report interface. This interface includes a map view, list view, and charts that invite comparison between incident report types and vicinity.

It was important to clearly show users whether information is verified. For simplicity, we use a simple checkbox to filter out all reports that are "unverified" or "verified false", even though we distinguish between these in the verification platform. We showed the verification status prominently on all report listings.

Reporters can select a neighborhood and choose a level of geographic precision. To protect the privacy of reporters, the system does not require reports to submit precise location information. Trusted administrators use the precise location when it is provided to verify reports. Regardless of the location precision, the public platform only shows reports to the neighborhood level.

Join our talk on Wednesday morning to hear more. We'll follow up later in the day with a Birds of a Feather session discussing how the open source data collection community continues to grow.

Hello San Francisco; Ready for FOSS4G

Mon, 03/09/2015 - 18:00



Landsat 8 image over San Francisco on December 31, 2014

We're out in San Francisco this week for FOSS4G North America. Look for Alireza and I at the conference or at our session Wednesday morning at 10:30am. Alireza will discuss using the OpenDataKit ecosystem to build mobile reporting and verification tools for refugee camps in Lebanon. We'll also be at the Birds of a Feather session discussing the future of OpenDataKit.

This week we will release new versions of landsat-util and Libra with some powerful new features. We are also working on some tools for managing OpenStreetMap data. We look forward to collaborating with old and new friends in the open source community to improve the tools for mobile data collection, Open Geo, and satellite imagery processing. Tweet us at @nas_smith or @scisco7 to meet up.

Tapping into OpenStreetMap Metadata

Thu, 02/19/2015 - 18:00



We just launched v0.1 of a new tool to tap into OpenStreetMap changeset metadata. We built the tool in partnership with the American Red Cross as part of the infrastructure for tracking efforts such as #MissingMaps.

OpenStreetMap changesets give us access to a wealth of metadata information that is not specifically geographic but incredibly rich. Metadata is helpful in understanding the changing nature of OSM. This is different from using geographic APIs like Overpass because metadata contains commit text, number of edits, which editor was used, etc. With metadata, we can track hashtags, analyze commit text or aggregate user metrics.

In 2014 alone, users committed over 6 million changesets to OSM. As OpenStreetMap's metadata grows, dealing with the sheer amount can be daunting. We built osm-meta-util as an experiment in making OSM metadata easier and faster to use.

osm-meta-util focuses on two core functions: downloading the minutely compressed metadata files and serializing into JSON. We convert compressed OSM XML files containing multiple commits to a stream of JSON objects that can be piped to any tool or API.

You can use the library in a Node application or as a command-line utility to download all the data between two dates:

MetaUtil({ 'start': process.argv[2], 'end': process.argv[3], 'delay': process.argv[4] }).pipe(process.stdout)

In combination with jq, to get a commit history we can simply run:

node app 001181708 001181721 1000 | jq -c '{date: .created_at, text: .comment}'

If you don't give the tool any parameters, it will get the latest changesets and update every minute.

We're using this utility to experiment building a metadata API with the American Red Cross. But we know there are many more uses of the rich OSM metadata and want to see what others can do with the tool. Together with the American Red Cross we've put this on OSM Lab, a Github organization for OSM related projects. Follow the development of osm-meta-util on Github.

Hello Mariano Arrien-Gomez!

Thu, 02/19/2015 - 11:00



Mariano Arrien-Gomez is joining Development Seed. Mariano is going to help us by creating beautiful illustrations and impactful visualizations. Mariano takes his inspiration from the world around him and expresses himself via illustration, photography and murals. Check out his Instagram feed for some examples.

Mariano is an artist who brings a meticulous and thoughtful approach to our design workflow. We're excited to see his work begin to influence our projects.

Say hello to Mariano on Twitter or Github.

Open Data Day Garage Party

Fri, 02/13/2015 - 10:00



We love open data. And we love to talk about it over drinks with other open data lovers. Next Friday, February 20th at 7:00pm we are hosting an Open Data Day celebration in the Mapbox Garage. Head over to the Garage after the first day of Open Data Day DC to talk about open tools for open data.

Even if you can't make it to Open Data Day DC (or didn't grab a spot on the now closed list), come over and share the #opendatalove.

Let us know if you're in by RSVPing now.

Data Hungry Happy Hour

Fri, 02/06/2015 - 10:00



We'll be at the Thought For Food Global Summit next week working with some brilliant people on the biggest challenges in agriculture and feeding the hungry. Look for Olaf at the Summit.

If you are in Lisbon for the Summit, please join us for a Data Hungry happy hour on Thursday evening. We'll be geeking out on better data, satellites, and sensors can contribute to better food policy. The Happy Hour kicks off at 18.30 in Fabulas in the center of Lisbon. The first couple of rounds are on us. You can RSVP here.

Welcome Mariano Arrien-Gomez

Mon, 02/02/2015 - 10:00



Mariano Arrien-Gomez is an artist and designer. He builds beautiful graphics and data visualizations that make our products more compelling, creative, and humane. Mariano directs his design skills toward the issues and topics he is passionate about, from preserving local parks, to genetically modified food, to soccer. He utilizes a range of visual mediums including illustration, painting murals, and photography.

In addition to his work at Development Seed, Mariano is active in the DC art scene. Mariano received a Bachelors of Fine Arts in Graphic Design from Virginia Commonwealth University. He is fluent in Spanish.

Announcing Libra - the Landsat imagery browser you will love

Thu, 01/22/2015 - 18:00



We've been working with Dauria Geo to produce the most usable imagery browser. Today we are releasing Libra, a fork of the Dauria Geo browser for open Landsat data. Libra allows you to browse, sort, and download more than 275 Terabytes of open Landsat imagery as easily as booking an Uber.

Liberating Landsat

We love open imagery. The global development organizations and developing governments that we work with use open satellite imagery for everything from evaluating disaster response, to tracking deforestation, to planning for drought. For our partners, open imagery isn't just a matter of cost; it is a matter of licensing and distribution. They get immediate access to Landsat images and can analyze, manipulate, and distribute with almost no restrictions.

To make Landsat data more useful, we've made it easier to use. We built two open source tools for working with Landsat data - Landsat-util and Landsat API. It used to take all day for Development Seed's imagery specialists to turn Landsat data into imagery layers for online maps. With these two tools, any developer can do it in a matter of minutes.

These tools gave us a huge head start in building Libra. Libra relies heavily on Landsat API to quickly query by date, geography, and cloud cover and get image URLs, scene centroids, scene boundaries, and other metadata. Using Landsat API as a backbone of Libra also encouraged us to make improvements and configuration changes on Landsat API such as changing the limits on requests and returned data and some error handling.

Have a look at Libra and hit us with your feedback @developmentseed.

Introducing Development Seed Lisbon

Tue, 01/20/2015 - 10:00



Last week we announced the opening of Development Seed Lisbon. To kickstart our operations in Europe, we brought on our friends from Flipside, an experienced team working on meaningful open data projects for organizations around Europe. This move allows us to connect with partners and talent in the region, and also deliver quality work right out of the gate.

Olaf Veerman

Olaf will lead the Lisbon office, run projects, and help us establish a strong presence in Europe. He lived for many years in Latin America, working with small business networks, cooperatives and small farmer groups in Brazil, Uruguay and Venezuela. His experience in working with civil society organizatons around the world allows him to quickly understand our partners' needs and help them use technology to increase their social and economic impact.

You can connect with Olaf through Twitter.

Daniel da Silva

Daniel brings solid engineering skills that he applies to anything from building light-weight frontends with well structured APIs, to deploying tools for offline/online data collection. He is a quick learner and problem solver whose technology expertise spans PHP, Node, Angular, Jekyll and Mongodb. Daniel is going to help us pick the right tool for the job and deliver quality work to our partners.

Find Daniel on Github.

Ricardo Mestre

Ricardo is a talented designer and front-end developer who pushes how modern technologies can be used to craft usable and engaging websites. He worked for some of the biggest companies in Portugal, but is most passionate about free culture and equality, which he contributes to through his music, art and other projects. Ricardo is going have a big impact on the design and usability of our work.

You can find Ricardo on Twitter.

Development Seed opens office in Lisbon

Thu, 01/15/2015 - 11:00



Development Seed is opening an office in Lisbon. Our team grows by a continent today in beautiful Portugal where we will continue to build data tools and solve complex development challenges. Establishing an office in Europe puts us closer to our partners in Europe, Middle East, and Africa. It will also allow us to better connect to the talented open data hacker movement in the region.

To bootstrap our European team we are immediately bringing on all our talented friends at Flipside. The Flipside team have been doing fantastic work on projects ranging from opening data on clean energy, building mobile monitoring tools with Text To Change, and tracking forest fires around Portugal. The entire team joins Development Seed today. Olaf Veerman from Flipside will lead the Portugal office and help us to grow the team.

If you are in Lisbon, come and celebrate with us tonight at our Open Data Happy Hour.

Building a Better Satellite Imagery Browser

Thu, 01/15/2015 - 11:00



Satellite imagery companies spend billions of dollars acquiring pictures of the planet. However, anyone who has worked with that imagery knows that the tools for searching and accessing this imagery are painful to work with. That's why we are excited to be working with Dauria Geo to build a fast, easy, imagery browser. To do that, we are relentlessly reducing friction and creating something that feels more like Airbnb or Pinterest than traditional GIS-based imagery browsers.

Dauria Geo's strategy is built on making imagery easy to integrate across every industry. Here is how we worked with them to bring usability to image browsing:


A primary goal is to eliminate the steps required to start looking at imagery. Dauria Geo discussed the "bounding box problem" -- traditional imagery browsers require you to provide a bounding box to start a search. The user either has to outline a polygon or upload a file with the bounding box. If the bounding box is too big or returns too many results these sites often require you to start over.

Our solution? Just zoom.

We will explore options for advanced search that may include boundary uploads, and those interactions will follow the same simple design concepts foundational to a good user experience. As we add those features, we will preserve the ability of users to immediately start interacting in a manner they expect from a modern website.

Responsive filter and sort

Once the user has identified their area of interest, the next step is sifting through all the returned images to find the best one for their needs. We've made this easier through visual filters and sorting tools that help you quickly get to the best image. Familiar icons allow users to filter and sort without having to learn the interface. We included visual tools like a histogram on our filters to help users intuitively understand the implications of their filtering choices. Sorting assumes "best to worst" to make sorting easy and to surface the best images. All along, we provide thumbnail previews of each image so that you can immediately see what you are getting.

Making the interactions obvious

Our aim is to create interactions that seem natural without the need for a website tour. Moving around the map shows coverage areas. Selecting an area filters results to that geography. We benefited here by using Landsat imagery that has a consistent bounding area, but we believe we can keep this intuitive even with different shaped scenes. To keep user interaction from breaking, we did a lot of work under the hood to make the site fast and responsive even while loading a lot of data and images. With a streamlined request process, once you've found the right image, you can download or purchase in one click.

We were able to quickly stand up a fast mapping application by building off the Mapbox stack. Simple, intuitive tools make it easier for more sectors to benefit from satellite imagery. As we help global development organizations and emerging economies to use imagery products, usable tools will be critical. We think existing imagery users will also appreciate having a more enjoyable relationship with their imagery finder.

Open Data Happy Hour in Lisbon

Fri, 01/09/2015 - 11:00



Next week we'll be in Lisbon where we're hosting an Open Data Happy Hour with our friends from Flipside. If you're in the neighborhood and want to talk Open Data, know more about our satellite imagery work, or geek out on sensors, make sure to drop by and have a drink with us.

The Happy Hour will be hosted at Liberdade 229 and start around 19:00. We hope to see you there. RSVP here.

Welcome Joe Flasher

Thu, 01/08/2015 - 10:00



Joe Flasher is joining Development Seed to help us run faster and to broaden our technology stack. Whether he's launching Delta II rockets, growing the open data community in Mongolia, or connecting an arcade claw game to a Facebook app, Joe is always looking for better ways to bend technology to solve real problems. He is constantly seeking the solution that is more elegant, more practical, and more human.

Joe is the perfect person to help Development Seed grow and explore. He's going push us to build better products and to make bigger open source contributions across satellite imagery, sensors, drones, and devices.

Give Joe a holler on Twitter or Github.

2014 Web Index measures the Internet's openness and impact

Fri, 12/12/2014 - 01:00


2014 Web Index measures the Internet's openness and impact

The World Wide Web Foundation released their 2014 Web Index, an annual accounting of how the internet has changed the economic, political, and social lives of citizens across every continent.

This is an ambitious and complicated story. The impact of the internet on the 84 countries in the 2014 report is difficult to isolate from culture, politics, and economics. Ultimately, the report found that open and inclusive internet correlate to equitable societies. The report is rich with data and worth reading in full.

Visualizing liquid data

Making useful and appropriate comparisons between 84 countries is hard. To create the Web Index, the Web Foundation pulled a range of primary and secondary data into normalized indicators, which together comprise the index. The report, produced online in collaboration with WESO, gives a global overview in addition to nuanced subviews on topics such as internet censorship and gender equality.

When we first began collaborating with Web Foundation to visualize this information, they were in the process of collecting and interpreting their data. They had identified key themes that they knew would be important: economic equality, neutrality, censorship and surveillance, and gender-based violence, among others.

Using these leads, we dived into the preliminary results in search of effective, telling visuals. Quick feedback loops and frequent communication with research specialists at Web Foundation helped us to build better visuals. It also helped Web Foundation better understand their data and how to communicate it.

Comparing Countries

We ran up against the challenge of showing 84 countries in interesting ways, while doing justice to the data. Although the data is country-level, not every story lends itself to a map. In our visuals, we use flags, bubbles, and names to indicate countries.

In plotting gender-based violence, we ran into the problem of page size. The graph examines countries with stark differences in how they support victims of online gender-based violence, and how frequently they prosecute perpetrators of that violence. Each graphic had to be single page and embeddable. To make this fit, we used a fisheye effect that moves with your cursor.

Open Data about the open Internet

The Web Index is fully open. You can download all of the data here. The Web Foundation and WESO built an open API with data from this year and previous years. The code for their site and our four visualizations are open as well.