Data Dissemination to the Government of Haiti

andrew — Fri, 05 Feb 2010 17:56:51 +0000

OpenStreetMap in the
World Bank Haiti Situation Room

For the past 3 weeks, since January 12, we've been actively involved in many efforts to provide support and analysis of the Haiti earthquake recovery. From immediate OpenStreetMap efforts to get imagery open licensed and traced, to working through CrisisCommons to get new volunteers mapping, creating dashboard common operating pictures of population affected by the event, emerging IDP camp locations, distribution centers, and gathering normalized health facilities around the country.

Historically, these tools have been useful for the public and media to cover an event - but the question often arose if they were beneficial to responders and citizens on the ground. It's been clear that tools such as OpenStreetMap have had a clear and positive impact on the response and recovery efforts. Organizations from the World Bank, to MapAction, to the Fairfax Search and Rescue teams have all been using OSM as a primary means of maps and routing. OSM volunteers have even been performing rough damage assessement and IDP camp identification.

While these tools have been useful for responders, there was still a disconnect on the availability of this data to the Haitian Government to access and assess the situations. The best data previously available before the earthquake is primarily from 2002, and at the same time most ministries are either non-existent or severely understaffed.

Bringing the Commons

Portable Hard drives with data
and map tools for the Haiti Government

In response, at CrisisCamp DC on Saturday the World Bank initiated an effort to provide the massive amounts of data and tools directly to the government. The goal was to provide a browser, like the Haiti Crisis Map with the imagery and OpenStreetMap roads, data gathering and visualization tools like GeoCommons, public and quality checked sources of Hospital locations, camps, and damage assessment, and the World Bank's own flyover imagery on portable hard drives and onto the desk of the Haitian ministers.

A major hurdle, however, was that there were no local copies of the raw imagery data. Thanks to the incredibly hard work of a number of organizations, especially Georgetown, SDSU and Internet2, we were able to move the 1.25 TB of raw data from the Hypercube server in San Diego to Georgetown's servers here in DC in a little over a day. These were loaded onto a series of 6 USB, self-powered hard drives. We chose the self-powered drives so it didn't depend on external plugs or fluctuating power sources.

In addition, we built an offline version of the Haiti Browser that can run by double-clicking an HTML file to run in a browser off of a hard drive or USB stick. You can get the code from here to see how it's done. We also included Delta State's MGRS Atlases for printing map books, and the GPS map images that can be used on Garmin handheld units.

There will also be several engineers deploying with the World Bank to assist in the dissemination of the data, working with the Haitian Ministries and President on using these tools and handling requests for more analysis and data with CNIGS (Centre National de l’Information Géo-Spatiale), Haiti's GIS department.

Enabling Self-Sufficient Government

The goal is to provide the Haitian government with their own capability of using the available tools for situational awareness and decision making. Beyond the immediate response and recovery efforts they are now enabled to utilize these tools for long-term reconstruction and infrastructure. The capabilities and tools should be sustainable and the goal is to close the data loop within the government as well as between external data collection and Haitian data needs and collection. We are moving beyond the traditional problems of leaving disasters without their own data or ability to be independent of the temporary organizations. The entire project is a model of how crowd-sourced data and tool development can have a beneficial impact both on remote support as well as directly to local citizens and government.

In the US and Europe, we're seeing a growing embrace of Open Government and transparency sharing data to citizens for engagement and collaboration. In Haiti, the community is performing a reverse-Gov2.0 - they're providing data and information to the government to respond, and rebuild their society.

A tremendous thank you to the numerous individuals and groups that helped and provided tools or data: World Bank, San Diego State University / Calit2, Internet2, Georgetown University, DigitalGlobe, Delta State University, Sahaha, Crisis Mappers, OpenStreetMap, NOAA, Ushahidi, DevelopmentSeed, TelaScience, STAR-TIDES, CrisisCommons, USAID, GeoCommons, OpenSGI, GeoEye.

Computer Friendly Spreadsheets for Sharing

Kate Chapman — Sat, 23 Jan 2010 05:52:17 +0000

One of the important aspects of sharing data is creating a format that is sharing friendly. A topic that frequently comes up at FortiusOne is bad spreadsheets. If you are going to share a spreadsheet there are a couple ways to make it easier for those involved. Developers usually think of spreadsheets are more of a database table, there are others that think of them as reports. Spreadsheets have a place in both situations. If it is meant to be a report, by all means make it a report. If you are recording data to share out make it more like a table. If people are going to mashup your data and input it into various systems the table version is much less of a headache. So what does it mean to make your spreadsheet table-like? You only have one set of columns in each page. Each column can have a header and probably should unless you are handing out metadata with your document, but sometimes metdata gets seperated from its data so that is not ideal. No columns should be merged anywhere in the page, each column should be the same width all the way down.

Hard to Ingest Spreadsheet

Easy to Ingest Spreadsheet