Hack for Nepal: A Media Lab Brainstorm

The 7.8-magnitude earthquake that struck Nepal on April 25 killed more than seven thousand people, injured twice as many, and left an estimated two million without homes. The United Nations reports that more than eight million people have been affected by the quake. Access to food, clean water, power, medical care, and other essentials are severely limited across the country. In the wake of such disasters, emergency relief must always be the first priority, and there are many organizations gathering and delivering resources for immediate aid. But while relief efforts in Nepal will continue for months and years to come, many aspects of the recovery process are either unmet, or could benefit from the tech savvy, creativity, and rapid prototyping that have become hallmarks of the hackathon.

On May 1, Nikhil Naik and Ramesh Raskar of the Media Lab’s Camera Culture group organized Hack for Nepal, a hackathon-inspired brainstorming session to discuss and share ideas for Nepal’s recovery. Two primary directives guided the day’s activities:

We encourage solutions that (i) are technology based, (ii) require minimal capital, and (iii) can scale easily as the number of users grow. (Examples: Safecast.jp for distributed radiation monitors; AirBnB home sharing.)
We discourage solutions that require systemic changes (changing the infrastructure), a lot of manpower (service heavy), or require heavy capital (best left to large companies). (Examples: changing government policy; paying many employees to run a service; deploying a new, unproven energy-generation solution.)
Special thanks to Lifelong Kindergarten student Srishti Sethi for setting up the Unhangout!

MIT students and Cambridge locals attended the event in person, but dozens more from around the world attended via Unhangout, a Media Lab tool based on Google Hangouts. On the Unhangout, participants broke into discussion groups in different rooms, and contributed ideas and links to resources on a communal working document that was also accessible to the in-person participants.

The event began with a presentation from Bigyan Bista, a PhD student in MIT’s Department of Biology and a member of MITeri, the Nepali Students’ Association at MIT. He began with an overview of the devastation and a summary of some of the most pressing needs that aren’t necessarily the focus of the major organizations:

  • Interactive mapping interfaces to allow “walkthroughs” of damaged areas
  • Utility to crowd-source the translation of rapidly changing websites in almost real time into Nepali
  • Disaster mapping beyond Kathmandu, namely in the districts of Rasuwa and Sindhupalchok
  • An inventory portal for each district to keep track of what supplies are abundant and what is lacking
  • A platform that focuses on water, sanitation, and hygiene
  • Internet/communications access in villages that have lost connectivity
  • Useful infographics for transparency in tracking aid

Bigyan then offered a list of organized relief efforts that are addressing some of the less obvious needs:

Coordination of relief efforts
EU resource dedicated to matching needs and supplies
Local mapping based on latest data
Near real-time situation report updates on social media
Coordination of volunteers on the ground via Facebook
One Stop Portal
New Initiative for Inventory portal

(All of the slides from Bigyan’s presentation are viewable here.)

The key, Bigyan noted, is not to duplicate the efforts of other organizations, but rather to focus on needs that aren’t being met. The most urgent need is for help with logistics, supply-chain management, and general coordination of relief efforts. Above all, Bigyan concluded, there are avenues for immediate, and also medium- and long-term engagement with recovery needs in Nepal.

The hackathon’s goal, therefore, was to identify problems and solutions, with an emphasis on rebuilding Nepal in the medium and long term.

Participants then broke out into discussion groups to focus on specific problems and goals.

Inventory, Aid Mapping, & Visualization

This group quickly identified a guiding concept: to create platforms for people to interact with disaster data in a way that is meaningful to the users and helpful to aid organizations. They envisioned a website and app with two functionalities–data collection and data visualization–with the goal of coordinating logistics more effectively via a centralized database. The group made a plan for a tool to centralize crowdsourced local data in machine-readable form, with update feeds to make data more actionable. Key components would include consistent format across data sources for more effective information processing; fields for machine readability and updates; geotagged pictures using mobile apps; and web forms to get information about needs in different locations.

Imagery & Mapping

This group discussed developing disaster maps using satellite and street-level data, and using the many video streams coming in to provide real-time data. They focused on addressing the greatest current need in this effort: mapping in remote areas, and in areas where damage is so extreme that they are unreachable from the outside. The Government of Nepal has created a portal to integrate various inputs and efforts; the group discussed ways of more efficiently aggregating data from various sources from satellites to cell phones, timestamping photos automatically to create a constantly updated “before/after” flow, and using non-image data, such as text messages, as inputs for mapping models.

In addition to this work, Nikhil Naik is leading a project to create a map visualization that will consist of a large number of geotagged images from before and after the earthquake.

Mobile Applications & Social Networks

This group brainstormed ideas for leveraging existing resources through apps and social networks. They discussed deploying mobile-device coverage through satellite-based connections; some cell towers might be down, and repairing them may not be a priority right now, but cell access is essential to coordinate aid efforts. They also considered deploying tools like Google’s Project Loon (balloon-powered Internet) and devices like BRCK, Ushahidi’s rugged, self-powered, mobile WiFi device.

Devices & Health

This group considered various issues related to health and sanitation, both in terms of immediate problems after the quake and more long-term problems and solutions. Access to potable water was a topic of particular concern. With the monsoon season about to begin in Nepal, and with many communities digging holes to serve as toilets, the group identified several priorities for quick hacks: geolocation of water sources and water purification equipment; adding lime to latrines to prevent disease; and distribution (web-based and printed) of basic public-health information on such topics as water safety, how to make toilets properly, and malaria prevention. Beyond water concerns, the group also discussed delivering basic onsite health training; using geolocation to post specific medical needs in remote areas; and delivery of medical supplies via drones.

Moving Forward

Since the hackathon, several self-organized groups have formed to take the ideas forward. The working Google document, generated organically during the event, has grown into a valuable online resource for the broader community interested in contributing to the relief and rebuilding process. A follow-up meeting will be organized soon with the help of MITeri and the broader Nepali community in the Boston area. We expect a few important projects to emerge out of this effort.

To support Nikhil Naik’s visualization project (see “Imagery and Mapping” above), we are seeking images via crowdsourcing. If you have images from Nepal that you are willing to contribute (especially from before the earthquake), please send a link to Nikhil <naik@mit.edu>.

Nikhil Naik is a PhD student in the MIT Media Lab’s Camera Culture group. He works with with Professor Ramesh Raskar on computer vision and computational imaging.

Photos by John Werner