UN Global Pulse Camp 1.0 | MIT Center for Civic Media
Nadav is a researcher and PhD candidate at the MIT Media Lab. He is a member of the Human Dynamics research group and the MIT Center for Civic Media. His research revolves around the intersection between communication networks, social dynamics, and systems that learn. A strong emphasis is given to systems and architectures that empower the edges of the network, giving end users control and ownership over their information and supporting security and privacy features., Prior to the Media Lab, Nadav had been working at a start-up company making communication chips for fiber-to-the-home networking. His R&D roles included system engineering and algorithm development. In addition, he performed consulting for several firms in areas of system engineering, product definition, network security, and locating seed investments. He received his bachelor's degree in electrical engineering from the Technion - Israel Institute of Technology cum laude in 2004, specializing in computer networks, machine learning, and signal processing.
UN Global Pulse Camp 1.0
(Photo credit: Christopher Fabian of UNICEF & Global Pulse)
Just got back from the UN "Pulse Camp 1.0".
Global Pulse is a new and quite ambitious UN initiative "to improve evidence-based decision-making and close the information gap between the onset of a global crisis and the availability of actionable information to protect the vulnerable" (Full overview at http://www.unglobalpulse.org/about).
The Global Pulse platform aims to do things like helping with the detection of slowly developing crises, by pulling in data from many different sources and analyzing them collaboratively. The platform will also attempt to to connect many of the relevant stakeholders, and give them tools to collaborate and coordinate among teams around the world, cross countries, cross sectors. The project will include local "Pulse Labs", that will be on the ground at different countries, staffed by experts, who will be in direct contact with local government, organizations, and communities, and help define their needs as well as provide development support for different platform components, and data analysis expertise. The first Pulse Lab will be launched in Uganda. (For more on the requirements and initial thoughts regarding architecture, see http://www.unglobalpulse.org/blog/self-assembly-required-real-time-platf... and http://www.unglobalpulse.org/blog/global-pulse-open-architecture-sustain...)
Sounds ambitious? Yes it is.
Sounds risky? Yes it is.
Sounds vague? Yes it is, at least for now.
Trying to answer these open questions like what is the Global Pulse platform, who are the stakeholders, what are challenges for building it - technical and other, and how can this initiative be successfully implemented, were some of the main drivers of the first Pulse Camp. The camp was held in NY over the period of 3 days (Dec 1-3), followed by a 2-day weekend hackathon organized by Random Hacks of Kindness (RHOK) in which the goal was to prototype and experiment with some of ideas coming out of the Pulse Camp.
The Director of Global Pulse initiative, Robert Kirkpatrick, and the rest of the Global Pulse team are going into this with their eyes open. They realize the great challenges they face, and are trying to define the platform in several stages and split it to contained and manageable "chunks". The project will rely heavily on existing open source projects and other existing platforms as much as possible, an open design approach, and the idea that all of its outputs will be open sourced as well.
Pulse Camp attendees came from many different places - a lot of UN folks from all over, representatives from different countries and regions, open source developers, NGOs, media, researchers and academics, interaction designers, and many others involved with crisis mapping, disaster response, and the other topics related to the Global Pulse platform.
The bulk of the camp's three days were spent doing a guided design process, based somehow on "agile" methodology, and facilitated by the folks from Thoughtworks. This including things like defining user "personas", telling their story, and then generating usage scenarios and from them technical requirements. Even though at times the techie folks like myself felt we just want to dive into the system design and architecture, I can understand how this process is useful for flushing out many use cases and also seeing what salient ideas and requirements emerge from the different workgroups. Also, this was a good way to involve the less technical participants and extract technical requirements out of their ideas. Aside from that there were different shorter talks about various technologies and other existing projects.
Some highlights and cool ideas that were raised:
- I really like the Global Pulse logo!
- At times it seemed as if each person has a totally different view as to what this system is. Is it essentially a social network evolving around data and analysis? Is it a front-end app for the user that provides all needs from end-to-end? Is it a collection of converters between data formats to make it easy for people to import/export into their favorite tool? Is it a tool that common citizens can interface with or is it centered around an analyst? And what is an analyst? do they know how to code or do they just use existing report templates?
- Same for what is a "workspace" - is it a place where all of an organization's data / users / reports lives? Or maybe a workspace = project, that a user opens, imports data feeds and invites peers? Or is it the local workspace an individual user sees when they work with the system? Some discussions got a bit confused since folks were using the same terminology but had different things in mind..
- Regardless of what the platform turns out to be, the Global Pulse team wants to make sure there is an end-to-end tool-chain that is open source and will be free for any country or organization to set up for free. End-to-end = from data collection through analysis all the way to distributing results/recommendations.
- "Hunch Lab" - great idea that came out of the UX working group. A place where analysts can pose questions, hypotheses, and ideas for further investigation, and have interaction between different actors similar to stackoverflow.com. Would love to see something like this for general academic research and in other fields.
- It was very cool to see how from separate discussions and workgroups, people came up with very similar ideas. One thing that emerged multiple times is the idea of the user interface having the look and feel of an Integrated Development Environment (IDE) traditionally used by software developers, a-la Eclipse. Only this one would be tailored for data analysis and research (our work-group called it IRE - "Integrated Research Environment", which could totally be implemented as an Eclipse plug-in.
- As always privacy issues about information and data sharing came up. One of the challenges is how to create a system that merges both public data and private data (e.g. Health data owned by one government which they don't want to expose to the world, but still be able to analyze it in context of other datasets, or private company data, or IRB protected data, etc.). Robert Kirkpatrick mentioned a future Global Pulse workshop centered around privacy architecture coming up soon.
- This was just the first in a series of Pulse Camps, on the road to define the system and get more individuals and organizations on board.
- At the end of the workshop the UN Secretary General Ban Ki-moon and the Dep. Director of NASA showed up in person to support Global Pulse and open the RHOK Hackathon. The Sec. General made a "I'm not a rocket scientist, but..." joke, and emphasized the importance of open source software. At the end he even received a Pulse Camp t-shirt! (see image above)
This post is getting way too long so I think I'll cut it here. I'll follow up in another post with my 2 cents about how the platform could look like and other random thoughts.
For a full summary of each day see the Global Pulse blog entries for Dec 1-4 (http://www.unglobalpulse.org/blog/all). Unfortunately couldn't attend the hackathon, so looking forward to find out what came out of that!