Thursday, September 4, 2008

InSTEDD introduction in Asia

I usually watch my very bright and articulate colleagues post blogs, but it’s been a great few weeks, with a collection of significant public efforts, and some of the efforts need notice. Let’s try a few photos this time.

My last blog was related to the Myanmar cyclone and our work on Sahana translation into Burmese, and, in the background, a very little about the Sichuan earthquake in China. I need to catch up a bit and put some other recent events in context.

Here at InSTEDD we’ve done a few things over the past three months that would be worth describing in detail but this blog would be excessively long. I’ll keep it more brief by just mentioning that we’ve:

1. Had useful meetings with the Gates Foundation,

2. Given a nice presentation at the Pacific Health Forum,

3. Hosted an interesting dinner at the Pacific Health Forum on "Ethics in Information Dissemination within Low-Resource Environments" (with WHO, PATH, Gates, Veratect, Grameen Bank, and a dozen others),

4. Chaired a useful workshop at SciFoo Camp on “Discovering Emerging Infections”,

5. Had some good conversations with HealthMap at Harvard on integrating our complimentary toolsets,

6. Presented at Kofi Annan’s Global Humanitarian Forum on using collaborative tools for climate change monitoring,

7. Had some fascinating and helpful discussions about design with IDEO,

8. Written a White Paper for the Rockefeller Bellagio Conference on “Collaboration in Emerging Infection Epidemiology”,

9. Co-chaired a track at the Fortune Brainstorm Tech conference on “Technology for Peace

And more. It’s been an incredible summer. I’ll also mention only briefly that we received a lovely gift from the Chinese government for our work with the Yunnan Center for Disease Control after the earthquake.

Let’s stay more current in the discussions below, but any reader can feel free to ask me about anything above.

Just looking over the past three weeks:


Dennis Israelski and I just spent a week in Bangladesh with both Grameen (Led by Yunus, Nobel Peace Prize, 2006) and BRAC, the largest NGO in the world. You’ll see a selection of photos below. Starting with Grameen, Dennis and I met extensively with the director of Grameen Solutions, Kazi Islam, had lunch with the Grameen Solutions Board of Directors, and took a trip to rural central Bangladesh as guests within a neighborhood Grameen Bank microfinance meeting.

We also visited one of the Grameen Health Centers where disease monitoring is taking place by sending staff from these centers to a network of villages each day, then reporting the conditions in the villages to a doctor in this center.

Those who run the Health Centers are asking for GeoChat on a model very similar to what we intend for the MBDS cross-border sites.

These ladies are a part of the network developing for village disease reporting because they are, of course, already a network.

Grameen neighborhood microfinance meeting, three hours north of Dhaka (Dennis and I were the guests of honor):

Grameen Microfinance loan family in Tangail village:

Grameen Solutions, Grameen Tower, Dhaka:

Grameen Health Center entrance (woman in pink coming out). Counter is the Grameen pharmacy:

Grameen Health Center clinic room:

A woman from Grameen Health Center who goes to six villages every day for health reports. This is our GeoChat use-case:

Shafqat, the CTO for Grameen Solutions. Ed and Robert wrapped into one, plus a beard, cap, and thobe. Speaks gorgeous British English, native Bengali. Smart, cheerful, very impressive man. The smile was constant. We saw a lot of his work.

BRAC: BRAC is, to my understanding, the largest NGO in the world. They work generally on primary education, children’s health, and gender-based microfinance. They have recently started a school of public health on a model that is considered one of the best in the world.

Dennis and I met with the Dean of that School, Professor Mushtaque Chowdhury, for a discussion of an Innovation Lab linkage between that school and the Cambodian School of Public Health, the topic to be “collaborative public health epidemiology and informatics”, teaching our methods and our tools sustainably in a format that gains credibility and academic rigor.

Professor Chowdhury (who is also a full Professor of Epidemiology at Columbia in New York) liked the idea very much, as did the Cambodian National Institute of Public Health, overseers of the School of Public Health in Phnom Penh. We’re working now on a proposal to link the two, since BRAC wants our tools for their school, Cambodia wants the BRAC curriculum model and our tools, and we want to ensure sustainability for the collaboration meme in public health informatics. How better than to embed it carefully and responsibly in the educational systems?

MBDS: Dennis and I then joined Ed Jezierski in Phnom Penh for the Mekong Basin Disease Surveillance Network (MBDS) Regional Forum, the Coming Out party for our tools and methods within the medical and public health providers for six nations in Southeast Asia. Photos below. Dennis, as co-chair of Information and Communications Technology (ICT), spent a lot of time at the podium and leading ICT workshops. We were (and are) only technical advisors and NOT technology implementers (Very Bad Idea) but Dennis had a decent bit of responsibility regardless.

We also, at this meeting, hosted a mini-conference for the demonstration of our tools. It took place on a Tuesday evening from about 5pm to 8pm. Photos below. We had a remarkable turnout – 73 registered attendees – and we had a TON of questions, comments, people trying GeoChat submissions in real-time to the number we gave them, and lots of post-event crowds around each one of us.

Despite our groaning internally about the somewhat clumsy delivery process we used for our message, and more technical difficulties outside of our control than we really wanted to endure (broken cables, failed phones, loss of hotel internet halfway through, etc) we were apparently a pretty good story anyway. We have a page of requests for collaboration including WHO’s WIPRO office in Manila, the Ministry of Health in Laos, the Yunnan CDC in China, ProMED-MBDS, US-CDC in Asia, RAND Corporation, and a dozen others. We were hot.

There will be a trial of our stuff around a dedicated ICT meeting in Savanakhet, Lao, probably in early November. At that point we will have Stung Treng and Champassak already explored in September (I hope), and we’ll arrive a day or two early in November and try linking – for the meeting – Mukdahan (Thailand) and Savanekhet (Lao PDR), and Savanekhet and Quang Tri (Vietnam). They are already rather tech savvy and have been cooperating with each other so this should be a small step for a trial. That will give us trials of cross border links in Cambodia, Vietnam, Laos, and Thailand before the end of the year. Not bad.

Dennis co-chairing ICT. He did an outstanding job on short notice, gently encouraging participation. It was gratifying to watch and very carefully and respectfully done:

Ed choosing phones for our workshop:

The MBDS InSTEDD Coming Out Party – GeoChat in real time from the audience:

More attended than we expected. We’d had a little warning that day so we shifted the workshop venue to the plenary hall.
(note the GeoChat submission in progress in the foreground)

Post-event discussions – we were pretty popular:


GeoChat: Now available. This week it grew from weak as a kitten to perhaps the maturity of Baron Rojo, Ed’s very large puppy. It’s not a mature tool yet, but it’s clearly no longer just embryonic potential. We used it in Phnom Penh with strangers on the fly and it worked well. I used it at Burning Man this past weekend. We’ve been asked to have it available for hurricane season, and we’ve been asked to introduce it to the WHO regional office, WIPRO, in Manila for their field staff.

OK, good stuff. We’re getting out there effectively. This one hit the mark.

GeoChat, the day before the MBDS demo:

Riff, our collaborative decision support tool, works well, and just as designed within the internal architecture, but it needs, I think, deeper exposure in the field to optimize the user interface. I think, to start, I want to start using Riff daily beginning this coming week within my own office.

As it happens, we had a PERFECT Riff problem appear in India while we were at the MBDS meeting and - disappointingly - we all wound up discussing it on a series of disjointed emails around the world. Riff is designed to help with precisely this problem - teams of people around the world collaborating quickly around a news item with a number of bots and automated services to help users make informed decisions with context. Unfortunately we were only a couple of days from having it ready. We’re there now, but we missed that brief window. Rats.

We are also making sure GeoChat messages go into Riff seamlessly and show up on a map module that we can then annotate for GoogleEarth through Mesh4x. We will be very sure our creations gracefully understand one another.

In Riff we really need to transform a complex set of capabilities into a look that is clean, intelligent and intuitive and I'm not sure we're there as well as we could be. This may be a place to ask for help:

If there are bright designers out there who would like to help with some open-source humanitarian software UI design that we give away for free for use in Very Bad Places, please drop a note to or write me directly at I’m looking for clever, creative ideas for helping Riff be a better interface on top of the already superb capabilities built into the features and modules. This tool has to be effective in helping users collaborate when they're hot, filthy, exhausted, a little scared, badly overburdened, and responsible for lives.

Below is an older version of Riff (perhaps three weeks), one view. It looks quite a bit better than this now, but it’s still in process. Note that the function really is quite decent. Its only the interface that is not yet intuitive.

Mesh4x, though the most technical of our products and the one with no user interface, is probably our most successful beast to date. It’s now been built into the OpenROSA effort called JavaROSA and will be introduced in Tanzania within the next month as a part of OpenROSA / OpenMRS and will have genuine use in the field. To quote Mulan, “our baby’s all grown up and saving China” (overstating somewhat and getting the geography wrong).

Mesh4x does a lot of linking from one thing to another, is fluent in several important computer applications, and plays very well with other communications devices.

Mesh4x is also, curiously, the easiest story to tell: Bad disaster, lots of tents on the hillside where humanitarian staff are working, assessments happening everywhere and being saved in Excel or Access or GoogleEarth, so lots of people collecting lots of information in stovepipes. Very inefficient, and maybe unsafe.

So then (socially) we agree to share information and bits of interest, then (technically) we introduce Mesh4x. Now, with Mesh4x helping, I move my pushpin on Google Earth in my tent, your Excel spreadsheet in your tent changes, we’re all sharing information in crisis, it’s efficient, with less redundancy, better use of donor dollars, and it’s all over SMS with no internet, using freely available libraries. Easy story. Great utility. Interesting idea. Perfect.

RNA is a set of tools (modules) within Riff for analytics and they are better than I thought it would be at this point. It might be the best of the Riff module set and has surprised a few people with it’s accuracy both in diagnosis and in defining relationships.

RNA is a distinct tools set, even though within Riff, and so we look at it separately. It’s apparently going quite well, using a team led by Dr. Taha Kass-Hout and involving Nico in Argentina and some interns from Trinity College who turned out to be excellent (or well-led? Or both?). We’ll keep this going. You can read about it on Taha’s blogs.

Tracker: Tracker is our web aggregator for health and humanitarian technology stories. We kept finding such great stuff buried in obscure places that we thought we had to build something that would allow us to share the new discoveries and capabilities with the entire humanitarian and global public health communities. It will be up for us this coming week.

Tracker, I’m told by reliable people, will be informative, well-designed, appealing to look at and interact with, and effective to use. We’ll know very shortly.

I’m pleased to note that, on Tracker, we’ve been involved with the humanitarian blogging community, UN relief agency staff and former staff, news organizations, NGOs, and others making sure content and form are aimed well. The feedback has been very positive and a number of sites are watching for us. I suspect we’ll get known pretty quickly.

Here is a screenshot from two days ago. Looks very encouraging. Clean, smart, interesting. We’re populating the resource databases this weekend:

As a last thought, my special compliments to
the Clarius team working so closely with us in Argentina and in Cambodia. Daniel, Nico, Laura, Miguel, Luigi, and the rest are truly among the very finest coders, testers, QA staff, designers, and teachers I’ve ever run across. They’ve been doing consistently outstanding work, under great pressure, across a 15 hour time difference, and across really challenging communication systems, in at least five countries.

All of their work has been done with care, cheerfulness, competence, and flexibility and the result has been a stunning output, with four major efforts ready for release from a small team in about eight months, starting from zero. It's a remarkable accomplishment.

Let me here, publicly, convey gratitude and admiration to each of them from all of the staff at InSTEDD.


Thursday, May 8, 2008

InSTEDD and the Myanmar Cyclone

I don't write often but I wanted to let everyone know what InSTEDD is doing in the Myanmar Cyclone response.

1. We have a long-standing working relationship with the open-source disaster management tool called “Sahana”, developed and managed through a group in Sri Lanka. We received a request from them for help early yesterday after Sahana had, in turn, received a request from inside Myanmar. That request was asking for Sahana as the disaster response coordination software to use for the humanitarian response, starting immediately and probably continuing for several months. We agreed to help, of course, and we’ve installed Sahana on the same system we use for our own website at Rackspace and it’s up and working now.

2. We heard late last night that we needed to localize Sahana into Burmese so that we could engage local staff in the response coordination process. We have now located some translators (not enough) for the 3500 words and phrases, and built tools to help break up the phrases into manageable chunks and then weave them back together again. Those translators are within our Google partners, our Mekong Basin Disease Surveillance System contacts in Asia, Stanford, World Vision, and more. An excellent example of crowdsourcing and networking and all happening literally overnight.

3. We’re also talking with the IBM Crisis Response Team that are trying to get into Burma and making sure we have the ability to link the peripheral Sahana server they'll have in Myanmar to our central Sahana server here in the US. We have a request pending with Google engineers, working with our own dev team, to help use our Mesh4x sync tools (used to connect databases to databases and systems to systems) to link center-to-edge in Sahana, making sure bandwidth is optimized in Myanmar and those outside Myanmar have a large pipe for access onto our hosted server at Rackspace.

4. We’ve had a request to push our internal GeoChat tool out early, really before several important features are established, and we’ve agreed in the interests of time and urgency to let that happen, even with a little risk to our reputation. GeoChat allows SMS messages to appear interactively on a map in any system (GoogleEarth, Visual Earth, Google Maps, GeoFusion, ESRI, GeoRSS feeds, KML feeds) and has proven so useful, even in current form, that the teams in Myanmar consider it worthwhile to take this relatively immature version even if only from a safety standpoint. It still works, very well, but it's neither elegant nor robust. But it will soon be both.

We still need Burmese translators for localizing Sahana, so please drop a note into Contact Us if you think you might be able to help.

This is a remarkable disaster, made very complicated by the communications restrictions placed on international responders by the Myanmar government. We hope those restrictions ease soon or this is going to be a very difficult response with excess deaths that will eventually be quantified and published.

I’ll keep updates going here.


Monday, January 28, 2008

New Staff at InSTEDD

We have two new members of the InSTEDD team and I couldn't be more pleased. You'll see their bios appear on the Bio page but Olaf and Suzanne are just as remarkable as the rest of the team, have both been working around us for a while, and will fit right in as full-time members of the staff.

Olaf Conijn - a Dutch development star

Olaf is a stunning coder, considered "incredible" by our director of engineering, Ed Jezierski (and Ed himself was considered a "force" at Microsoft, leading the "Patterns and Practices" group before joining InSTEDD, so one star is apparently able to recognize another...) . Ed's known Olaf for a while and has watched him develop software for several years. Olaf, though, is modest about his achievements, noting only that, as a teenager, he found he "liked writing computer programs". Only when pressed does he mention that he was teaching coding to world-class professionals (some at Microsoft) when he was 17. He's now all of 25 and living in Amsterdam. He'll be working on platform development and tool integration for our deep-field cross-border reporting systems, and helping national Ministries of Health meet reporting requirements for the new International Health Regulations that came into effect in June of 2007.

Suzanne Jul - a perfect match for InSTEDD

Suzanne is a PhD computer scientist and disaster maven with a Masters in Computer Science from the University of Washington and a PhD in Human-Computer Interface from the University of Michigan. She then got interested in disaster response tools, became a Red Cross Disaster Services volunteer, took courses in disaster response and International Humanitarian Law while consulting for Cisco, worked in the Katrina response and in disaster exercises for the State of California, then showed up on our radar after working at the ISCRAM Conference in the Netherlands last year (Information Systems for Crisis Response and Management). I talked with her during a meeting at the Technical University in Delft and was impressed with both her career achievements and her pragmatic intelligence. Lots of common sense, and a huge desire to put all of that careful training to use somewhere that matters. So she started at InSTEDD last week. I've seen her portfolio and she's perfect for the interface design problems we've seen all over the humanitarian support space.

The team is looking at shortcomings in humanitarian tools

About that interface design problem. Suzanne is particularly important because so much (not all, but most) humanitarian software has been either designed for other tasks and adopted simply for ubiquity and access, or it was designed precisely to meet a perceived need by someone who DOES that job, but isn't trained in software design, quality assurance, or interface optimization. Now InSTEDD has the luxury of staff trained in all three, PLUS experience in the field and a near-constant exposure to real field conditions because we test our stuff in places like the Mekong Delta, Olympic National Park, and the Costa Rican cloud forests.

Nothing is ever guaranteed, but this is shaping up fairly well.


Seoul and Water Disasters

I’m in Seoul, Korea attending the UN’s “Second Meeting of the High-Level Expert Panel on Water and Disaster” and I’ve learned a great deal in just a few hours. Participants around me are from Bangladesh, England, UNICEF, the Japan Water Forum and a dozen others, and the opening speech this morning at 9am was delivered by a man who was, unexpectedly, named Prime Minister of South Korea at 10:30.

I’ve been uncharacteristically quiet today, listening to the problems seen in Cyclone Sidr in Bangladesh, and the arguments for including slow-onset events under the label “water disasters” (like groundwater depletion and river contamination in southern India, for example) so that they can get the attention they need. It's a small room at a nice hotel (made smaller by a LOT of media cameras) and the participants seem to know each other fairly well. The conversation has been structured, but informal, and I've noted the remarkable range of problems we face globally that are rarely considered in the West. After one long description of the consequences of livestock loss in hurricanes, there was an extended discussion from Colombia, for example, on how to make Disaster Risk Assessments as much a part of development and reconstruction as an Environmental Impact Statement. And how to make disaster risk reduction as ubiquitous an effort within a population as seatbelts are for injury avoidance and safe sex for HIV transmission reduction.

I spoke on the place of InSTEDD and our partners in the aftermath of acute water disasters like floods and typhoons, including some of the tools we’ve found or built that might be helpful there. The talk was apparently very well-received and we’ve a new set of conversations underway with colleagues we’ve just met this morning. One discussion included a nice idea for linking a regional disaster response planning team in Central America using VSee (

After my talk (which was on requirements for effective response to disasters in general, not just disease reporting) I heard that peasants in Colombia, a country riven by violence from narco-trafficking, still consider floods a greater threat than anything else in the nation. And that India now gets 50% of her daily water from underground aquifers - water that’s been percolating slowly downward for millennia, is now being withdrawn at a far greater rate than it’s being replenished, and is clearly a disaster in the making.

One of the most interesting items I heard was that the Netherlands has a very clever initiative for improving global water and sanitation called AKVO ( They are using tools based on Wikipedia, Ebay, and RSS to help link local communities, trusted local partners, lenders, and donors in a mesh of small-scale funding that is directed to clean water and sanitation provisioning. It's efficient, and transparent, and a decent model for linking those who have to those who need. It’s a nice idea and I wish them great success.

My presence at this conference, requested by a respected colleague from the Canadian Foreign Ministry, is not directly related to the immediate efforts at InSTEDD, but the attendees apparently found what I talked about worth discussion and inclusion, and I’ve now learned a lot about topics I’ve never examined before. It’s just an extra overnight stay as I pass through Seoul since I’m on my way to Bangkok to join the team in Thailand for a few days.

We'll meet up in Thailand and, after talking with some members of the Mekong Basin Diseases Surveillance consortium in Cheng Mai later this week, we'll migrate into Cambodia for some deep-field epidemiology assessments near the Lao border and a large set of conversations with people and organizations already working in the region.

I'll leave Seoul tomorrow, but if anyone has thoughts regarding support to clean water and sanitation needs after disasters, please drop a note into the forums and we'll make sure your ideas get to the conference coordinators. Be sure to tell us if it's ok to have them contact you and how they should do it.


Sunday, January 13, 2008

Looking around at the beginning

I've had an interesting and exciting few months since arriving as CEO at InSTEDD in October. Much of that time has been spent traveling with colleagues to visit difficult areas around the world, talking with professionals expert in epidemiology and disaster response, and hearing from communities and governments how hard it is to watch for diseases and prepare for disasters in the face of innumerable other obligations in their days. Through subsequent conversations with the superb staff at InSTEDD, we've been sorting out how and where InSTEDD can be most effective and we have some ideas.

We're now designing tools and developing partnerships that we think might help us put together resources that make sense. Over the next few months I'll mention a few of those capabilities we're thinking about, and why, and I'll ask for your opinions. When we later find something interesting and useful we'll talk about it openly, evaluate the pros and cons honestly, try what seems useful in the field, and tell you here, clearly and fairly, how it went.

When it comes right down to it, those on staff at InSTEDD have quite a bit of experience, so know very well how big the world is, how many smart people are already trying to help around the world, and how little we know about what's out there. So we'll be obvious when we ask for advice, and I look forward to hearing from those who can help us design and build simple, robust, effective, and free tools for the humanitarian community.