Tag Archives: UNICEF

Automatically Classifying Text Messages (SMS) for Disaster Response

Humanitarian organizations like the UN and Red Cross often face a deluge of social media data when disasters strike areas with a large digital footprint. This explains why my team and I have been working on AIDR (Artificial Intelligence for Disaster Response), a free and open source platform to automatically classify tweets in real-time. Given that the vast majority of the world’s population does not tweet, we’ve teamed up with UNICEF’s Innovation Team to extend our AIDR platform so users can also automatically classify streaming SMS.

BulkSMS_graphic

After the Haiti Earthquake in 2010, the main mobile network operator there (Digicel) offered to sent an SMS to each of their 1.4 million subscribers (at the time) to accelerate our disaster needs assessment efforts. We politely declined since we didn’t have any automated (or even semi-automated way) of analyzing incoming text messages. With AIDR, however, we should (theoretically) be able to classify some 1.8 million SMS’s (and tweets) per hour. Enabling humanitarian organizations to make sense of “Big Data” generated by affected communities is obviously key for two-way communication with said communities during disasters, hence our work at QCRI on “Computing for Good”.

AIDR/SMS applications are certainly not limited to disaster response. In fact, we plan to pilot the AIDR/SMS platform for a public health project with our UNICEF partners in Zambia next month and with other partners in early 2015. While still experimental, I hope the platform will eventually be robust enough for use in response to major disasters; allowing humanitarian organizations to poll affected communities and to make sense of resulting needs in near real-time, for example. Millions of text messages could be automatically classified according to the Cluster System, for example, and the results communicated back to local communities via community radio stations, as described here.

These are still very early days, of course, but I’m typically an eternal optimist, so I hope that our research and pilots do show promising results. Either way, we’ll be sure to share the full outcome of said pilots publicly so that others can benefit from our work and findings. In the meantime, if your organization is interested in piloting and learning with us, then feel free to get in touch.

bio

WHO Using UAVs to Transport Medical Supplies (Updated)

Update: DHL to deliver medicine via UAV [link]

The World Health Organization (WHO) is experimenting with Matternet’s new quadcopters (one of which is pictured below) to transport medical supplies to remote regions in Bhutan. The country lies in the Himalayas, which makes access to public health particularly challenging for rural communities. Reaching these remote mountain populations in a timely and affordable way is key. This explains why WHO is looking into UAVs.

Screen Shot 2014-08-26 at 3.44.21 PM

Matternet is “aiming to build a network of low-cost quadcopters to connect the country’s main hospitals with rural communities.” The team “uses small quad- copters that can carry loads of about four pounds across 20 km at a time, to and from pre-designated landing stations. The company is able to track these flights in real-time, and aims to eventually deploy fully-automated landing stations that replace drone batteries, giving them extended range and flight time. The drones it uses typically cost between $2,000-5,000.”

WHO UAVs

WHO is not the only international humanitarian organization exploring the use of UAVs for the transportation of small payloads. Colleagues at UNICEF and Médecins Sans Frontières (MSF) are actively exploring this use-case as well with the latter in early pilot stages with Matternet in Papua New Guinea.

Screen Shot 2014-08-26 at 5.46.28 PM

UAVs can also be used in other ways to support public health projects. Take my UAV colleagues in the Philippines who are collaborating with the United Nations Development Program (UNDP) on a food security problem with obvious linkages to public health. Typhoon Haiyan uprooted millions of coconut trees when it barreled through the country. Many of these trees have since been rotting, which is now leading to a Rhinocerous Beetle infestation that can wipe out the entire coconut industry—a very important source of livelihood for many in the country. Meanwhile, other colleagues in Pakistan are looking into using UAVs “to identify and exterminate dengue larvae” as part of an existing intervention that uses smart phones to promote mosquito mitigation efforts.

bio

See Also:

  • Humanitarian UAV Network: Strategy for 2014-2015 [link]
  • Humanitarians in the Sky: Using UAVs for Disaster Response [link]
  • Crisis Map of UAV Videos for Disaster Response [link]
  • Humanitarian UAV Missions During Balkan Floods [link]
  • “TripAdvisor” for International UAV/Drone Travel [link]
  • UAVs, Community Mapping & Disaster Risk Reduction in Haiti [link]

Could Social Media Have Prevented the Largest Mass Poisoning of a Population in History?

I just finished reading a phenomenal book. Resilience: Why Things Bounce Back, was co-authored by my good friend Andrew Zolli of PopTech fame and his won-derful colleague Ann Marie Healey. I could easily write several dozen blog posts on this brilliant book. Consider this the first of possibly many more posts to follow. Some will summarize and highlight insights that really resonated with me while others like the one below will use the book as a spring board to explore related questions and themes.

In one of the many interesting case studies that Andrew and Ann discuss in their book, the following one may very well be the biggest #FAIL in all of development history. The vast majority of Bangladeshis did not have access to clean water during the early 1970s, which contributed to numerous diseases that claimed hundreds of thousands of lives every year. So UNICEF launched a “nationwide program to sink shallow tube wells across the country. Once a small hand pump was installed to the top of the tube, clean water rose quickly to the surface.”

By the end of the 1970s, over 300,000 tube wells had been installed and some 10 million more went into operation by the late 1990s. With access to clean water, the child mortality rate dropped by more than half, from 24% to less than 10%. UNICEF’s solution was thus “touted as a model for South Asia and the world.” In the early 1980s, however, signs of widespread arsenic poising began to appear across the country. “UNICEF had mistaken deep water for clean water and never tested its tube wells for this poison.” WHO soon predicted that “one in a hundred Bangladeshis drinking from the contaminated wells would die from an arsenic-related cancer.” The government estimated that about half of the 10 million wells were contaminated. A few years later, WHO announced that Bangladesh was “facing the largest mass poisoning of a population in history.”

In a typical move that proves James Scott’s thesis Seeing Like a State: How Certain Schemes to Improve the Human Condition Have Failed, the Bangladeshi government partnered with the World Bank to paint the sprout of each well red if the water was contaminated and green if safe to drink. Five years and over $40 million later, the project had only been able to test half of the 10 million wells. “Officially, this intervention was hailed as almost instantaneous success.” But the widespread negative socio-economic impact and community-based conflicts that resulted from this one-off, top-down intervention calls into question the purported success of this intervention.

As Andrew and Ann explain, water use in Bangladesh (like many other countries) starts and ends with women and girls. “They are the ones who will determine if a switch to a green well is warranted because they are the ones who fetch the water in water numerous times a day.” The location of these green wells will largely determine “whether or not women and girls can access them in a way that is deemed socially appropriate.” As was the case with many of these wells, “the religious and cultural norms impeded a successful switch.”

In addition, “negotiating use of someone else’s green well was an act fraught with potential conflict.” As a result, some still used water from red-painted wells. In fact, “reports started to come in of families and communities chipping away at the red paint on their wells,” with some even repainting theirs with green. Such was the stigma of being a family linked to a red well. Indeed, “young girls living within the vicinity of contaminated wells [recall that there were an estimated 5 million such wells] suffered from diminishing marriage prospects, if they were able to marry at all.” In addition, because the government was unable to provide alternative sources of clean water for half of the communities with a red well, “many women and girls returned to surface water sources like ponds and lakes, significantly more likely to be contaminated with fecal pathogens.” As a result, “researchers estimated that abandonment of shallow tube wells increased a household’s risk of diarrheal disease by 20%.”

In 2009, a water quality survey carried out by the government found that “approximately 20 million people were still being exposed to excessive quantities of arsenic.” And so, “while the experts and politicians discuss how to find a solution for the unintended consequences of the intervention, the people of Bangladesh continue bringing their buckets to the wells while crossing their fingers behind their backs.”

I have several questions (and will omit the ones that start with WTF?). Could social media have mitigated this catastrophic disaster? It took an entire decade for UNICEF and the Bangladeshi government to admit that massive arsenic poisoning was taking place. And even then, when UNICEF finally responded to the crisis in 1998, they said “We are wedded to safe water, not tube wells, but at this time tube wells remain a good, affordable idea and our program will go on.” By then it was too late anyway since arsenic in the wells had “found their way into the food supply. Rice irrigated with the tube wells was found to contain more than nine times the normal amount of arsenic. Rice concentrated the poison, even if one managed to avoid drinking contaminated well water, concentrated amounts would just up in one’s food.”

Could social media—had they existed in the 1980s—been used to support the early findings published by local scientists 15 years before UNICEF publicly recognized (but still ignored) the crisis? Could scientists and activists have launched a public social media campaign to name and shame? Could hundreds of pictures posted on Flickr and videos uploaded to YouTube made a difference by directly revealing the awful human consequences of arsenic poisoning?

Could an Ushahidi platform powered by FrontlineSMS have been used to create a crowdsourced complaints mechanism? Could digital humanitarian volunteers from the Standby Volunteer Task Force (SBTF) have worked with local counterparts to create a live country-wide map of concerns posted anonymously by girls and women across thousands of communities in Bangladesh? Could an interactive voice response (IVR) system like this one been set up to address concerns and needs of illiterate individuals? Could a PeaceTXT approach have been used to catalyze behavior change? Can these technologies build more resilient societies that allow them to bounce back from crises like these?

And since mass arsenic poisoning is still happening in Bangladesh today, 40 years after UNICEF’s first intervention, are initiatives like the ones described above being tried at all?