Grameen’s AppLab comes of age

Today is a very exciting day for many colleagues in Uganda, a day which sees the launch of a suite of new services from Grameen’s AppLab project. I was fortunate enough to be involved in the very early stages of the initiative, spending a month on the ground studying a mixture of geography, culture, challenges, data availability and technologies in and around Kampala (and occasionally beyond).

One of the best times to be involved in something like this is at the very beginning – the time when everything is on the table, nothing is ruled out and there’s no such thing as a bad idea. Over the course of the month we came out with around fifty ideas for mobile services, based on our research of the Ugandan landscape, and the kinds of issues, gaps and concerns which potentially lend themselves to a mobile solution.

A large part of the fun is without doubt this multi-faceted research – understanding the landscape from multiple perspectives and sources. TV, radio, conversations with taxi drivers (who, regardless of where they drive seem to have answers to all the world’s problems), newspapers, villagers, village phone operators, waiters, children and eavesdropping conversations in bars, all of which helps build a picture of what matters to people and what doesn’t.

Vision, Uganda
Image: Understanding local and national issues is an essential starting point in the mobile applications development process

Although it’s vital to start with the need, figuring out how to meet it becomes the next big challenge. Rural communities aren’t just passive recipients of information, but content generators in their own right. Communities are rich with knowledge, but more often than not this knowledge – not to mention more official sources of information – are rarely stored in anything resembling digital-friendly. Finding out who has the information you need, who owns it, how often it gets updated and how it’s stored are all part of the ongoing puzzle.

One of the most interesting and exciting phases of the AppLab work was the rapid protoyping – getting out into the field (or the matatu [bus] stations, to be precise) and offering people the opportunity to text in agriculture- or health-based questions. Any questions. What seemed to them like a smart, fully-automated system was in fact a handful of health and agriculture students sitting at computers in the MTN/AppLab offices, manually reading incoming questions and formulating 160-character answers. Suffice to say, the data gathered over a few days gave the strongest indication yet of the need and perception of such a service to potential users. The value of this kind of work cannot be understated.

Rapid Prototying (Photo: AppLab)
Photo: Students respond to incoming queries using the early version of FrontlineSMS, which was set up to help gather the data

Going back to today’s announcement, out of the original fifty early-stage ideas, AppLab have launched an initial suite of five:

Health Tips
Provides sexual and reproductive health information, paired with Clinic Finder…

Clinic Finder
Helps locate nearby health clinics and their services

Farmer’s Friend
A searchable database with both agricultural advice and targeted weather forecasts

Photo courtesy AppLab

Google Trader
Matches buyers and sellers of agricultural produce and commodities as well as other products (Google explains how it works here)

As part of the initial research, we looked at a whole suite of technologies on which to base solutions, including J2ME, WAP, high-end smart phones, 3G and MMS. As is usually the case, however, SMS won through and all of the services launched today are, according to AppLab,  SMS-based and:

designed to work with basic mobile phones to reach the broadest possible audience. Users can access the services quickly and privately at the time of their choosing and search relevant content on-demand, like someone with access to the Internet

A lot of work continues to go into AppLab’s work in Uganda, and today hopefully marks the beginning of many new announcements (believe me, many other exciting initiatives are already in pilot stage). By working through existing structures in the country (principally MTN and the Grameen Village Phone network, not to forget Google’s growing influence), AppLab is well-placed to identify, build and deliver appropriate, relevant mobile-related services to local communities, and my congratulations go out to David, Eric and everyone who has worked so hard on the project over the past two years.

For a little more indepth analysis on today’s announcement, check out White African’s excellent blog post and the short Grameen video below. The official Press Release is available here.

Coffee, Clark, Careers

All great journalists immediately put you at ease. Clark Boyd, someone I’ve been extremely fortunate to have spoken to on a number of occasions, is one of them. Interviews feel more like chats over cups of coffee in the dentists waiting room than recorded interviews set to go out over the airways in the US (and beyond).

Clark recently got in touch and asked if I’d be interested in giving a little careers advice – not to him but to people interested in mobile, technology, Africa and so on. Never one to turn down the opportunity, we recently sat down for coffee at my village dentist and chatted over coffee. Since these are the kinds of questions I regularly get asked by students and others interested in my work, it seemed sensible to re-post it. So, here you go. Apologies for Clark’s choice of picture. The original post is here.

Ken Banks: Cell Phones on the Frontlines

Ken Banks, kiwanja.netI have to say, this Wide Angle assignment was a tough one. In my nearly 6 years of covering technology now, I have to say I’ve come across quite a few people who have very, very cool jobs. But few people with those cool jobs have the drive, energy and determination that the man at right does. This is Ken Banks, and his online home is The tagline for the site says it all: “where technology meets anthropology, conservation and the development.” Ken is as close to a true “renaissance man” that I’ve come across in my forays into technology across the globe. His interests seem as wide and varied as his abilities. And the fact that he’s managed to somehow combine those interests and abilities into a career is, even to this jaded journalist, inspiring.

I’ve done stories on a number of Ken’s efforts in the past few years. The one that really grabbed my attention is a project Ken’s been working on called FrontlineSMS. So, for this Wide Angle Podcast, I begin by asking Ken to describe FrontlineSMS in his own words.

(Clark’s podcast can now be found on the kiwanja “Audio/Video” page here)

(Picture comes courtesy of Ken’s friend, another guy with a cool tech job, Erik Hersman)


As the latest iteration of FrontlineSMS celebrates its first birthday, now is as good-a-time as any to sit back with a double latte and reflect on the successes – and challenges – of the past twelve months

Happy birthday to \o/

Exactly one year ago today – 25th June, 2008 – saw the release of the revised version of FrontlineSMS. Originally written in 2005 over a five week summer break in Finland, the MacArthur Foundation stepped in a couple of years later to fund the development of an improved, Java-based multi-platform version following its successful deployment in the Nigerian Presidential elections, an event which turned out to be pivotal in the short history of the software.

FrontlineSMS election monitoring

Since then there have been numerous interim releases thanks to follow-on funding from the Open Society Institute (OSI) and the Hewlett Foundation, funding which has allowed us the luxury of pushing out new and improved features to users. In three months time we’ll see the release of a significant upgrade with a range of new features – including picture messaging – which promises to open up the platform yet further.

Suffice to say it’s been an exciting if hectic past twelve months. From a standing start last summer almost 2,600 NGOs have – one way or another – got to hear about the software and downloaded it. The online community is picking up steam and currently stands at just shy of 600 members. We’re about to launch another community, this time aimed at developers, although thankfully we won’t be starting from scratch on this one. Numerous programmers, projects and organisations have already begun working with the code, and others have used it to build diverse applications including human rights- and agriculture-based SMS solutions. We’re very excited where this could be headed, but hold no illusions that an active, engaged developer community is essential if FrontlineSMS is to ever become sustainable, developmentally at least.

Over the past twelve months the software has also been profiled in the international media and presented at dozens of ICT4D and non-ICT4D conferences around the world. Thanks to the efforts of the FrontlineSMS:Medic team it also made it onto the CNN and Discovery Channel websites (a selection of videos are available via the community site). Recent efforts by the University of Canberra have also attracted the attention of Brad Pit and Angelina Jolie.

FrontlineSMS Chinese

In addition, volunteers have helped translate the software into over half-a-dozen languages, numerous non-profit organisations have held FrontlineSMS training sessions for their staff, and Plan International went a step further and made their own introductory videos to distribute to Plan field offices via DVD. The software has a growing following on Facebook and Twitter, and continues to get good traction in wider non-profit circles. Many people are beginning to hear about FrontlineSMS through their own networks, and that’s a simple yet significant first step for many NGOs looking to deploy SMS services in their social change work.

But the past year – and the three before that, if we’re to be honest – have presented their fair share of challenges. When FrontlineSMS was originally conceived in 2005 it was highly experimental. Every step since has presented its own mini mountain to climb.

#1 Applications development and deployment

We’ve certainly learnt a lot on the development side of things – everything from the technical, geographical and financial challenges of creating mobile tools for use in resource-constrained environments, through to ways of encouraging adoption among the hard-to-reach grassroots NGO community. Best practice – such as understanding the fundamentals of the problem at the very beginning, researching existing solutions, rapid prototyping, choosing appropriate platforms and leveraging existing structures, are among a number of important key steps we’ve learnt along the way.

#2 Dismissing myths

By far the best way to fully understand the geographical, cultural and technical challenges of building mobile tools is to go out and build tools. It’s also imperative to spend a little time with the people or organisations you’re trying to help, ideally before you start designing anything. Neither of these are easy, mind you – time and cost are two factors – but you’ll gain far greater insight than you could ever do basing your project entirely on what other people say. Misconceptions and hype can be dangerous distractions and blind alleys, particularly if you base some of the fundamentals of your project on them. Scaling, “big is beautiful”, centralised versus distributed and “collaboration at all costs” are just a few. Proceed with caution.

#3 Fragmentation

Although things are much better than they were with the 2005 release (which only supported half-a-dozen devices), today we face the challenges created by a hugely fragmented handset market which innovates and releases new models at a remarkable rate.


We could, of course, make things easy for ourselves and insist users get hold of a GSM modem, but the whole point of FrontlineSMS is to try and lower the barrier to entry, and only supporting modems (and not the devices most readily available to NGOs – mobile phones) would be entirely counter-productive. Over the coming weeks we hope to gain access to test facilities which will allow us to test the software on a wider range of more recently-released phones, but keeping up with the industry will be an ongoing challenge.

#4 Device configuration

Getting their phone connected and working remains one of the bigger challenges for end users. We’ve made things as easy as possible with built-in automatic detection and configuration, but this can be totally thrown by cheap or fake/imitation cables, handsets not put into the right ‘mode’, incorrect or corrupt software drivers, software running on the computer ‘locking’ the phone and not allowing FrontlineSMS access, or by the phone communicating using different ‘non-standard’ protocols. Again, this is a problem which won’t go away, but one we’re going to have to do our best with.

#5 Outreach

Finally, how do you promote a grassroots messaging tool to grassroots NGOs who by their very nature are largely ‘offline’? It may have taken four years, but FrontlineSMS is today relatively well known among our target communities, and much of the adoption is driven organically, i.e. by one NGO talking about it with another. We’re very happy with our 2,600 NGOs in the past year, but we should be reaching out to many more. Getting it up to nearer 5,000 over the next twelve months is the target we’ve set ourselves.

#6 Bandwidth

FrontlineSMS is run by the smallest of teams – one project manager and two part-time software developers. As the software begins to get serious traction in a number of target areas, we need to plan very carefully to ensure that we don’t miss out on key opportunities, that we keep up with the users’ needs, that we remain responsive, and that we don’t become too overstretched. Some of the signs tell us that perhaps we already are.

#7 Evaluation and impact

A number of researchers and larger international NGOs are beginning to take an interest in helping us assess the impact of FrontlineSMS among the user base. There is still much we don’t really know – how many downloads lead to regular use (and how many don’t), what barriers stop people using it, what impact the software has on the work of the NGO, and the benefits it brings to the communities they serve. These are all critical questions.

FrontlineSMS around the world

As we enter the second year, we’re just about ready to start making sense of the map and what it represents. And that means our successes, our failures, and where we go from here.

FrontlineSMS reaches out in Cambodia

Courtesy Bendigo Advertiser

Brad, Angelina team up with FrontlineSMS project in Cambodia:


Researchers from the University of Canberra are working with the Maddox Jolie-Pitt Foundation (MJP), an organisation founded by philanthropists Angelina Jolie and Brad Pitt, to trial a cost effective text-messaging system. The project with the University of New England is funded by the Australian Centre for International Agricultural Research to help improve the lives of some of the poorest people in Cambodia.

The mobile phone system can be used to alert villagers about disease outbreaks, like tuberculosis and other important health and agricultural information, like pest problems in Samlaut, a remote and rural outpost in northwest Cambodia where MJP is implementing large scale integrated development interventions.

Although reports are showing that the percentage of Cambodia’s population falling ill with TB declined in 2007, Stephan Bognar, Executive Director of MJP said that Cambodia had approximately 495 new cases per every 100,000. Only Zimbabwe and South Africa had higher rates. “We need, therefore, to find effective tools to quickly disseminate information to isolated and rural communities.”

The University’s Dr Robert Fitzgerald says the researchers chose SMS text messaging as the information tool of choice because Cambodia has 85 per cent mobile phone coverage and texting can cost as little as 3 US cents.

The seed for the simple, but successful technology came from a project to help Cambodian farmers get better prices for their crops.

“We wanted to help farmers access the price of maize or soybeans on demand, so they were in a stronger position to negotiate the sale of their crop,” Dr Fitzgerald said. “Many of these farmers have to borrow money to plant their crop, in most cases paying 4 per cent interest per month.”

“Some traders and farmers knew some price information but that was not always shared, so the only way the farmer could find out about the price was to travel to town or ask nearby friends. So now farmers can text the agricultural information server and straight away find out about the prices.”

NAMA FrontlineSMS server: Robert Fitzgerald with Moul Sam who runs the server (Photo: Rob Fitzgerald)

The FrontlineSMS system runs on any PC, with a GSM modem and a pre-paid SIM card. The software is free and open source, and was developed by UK-based expert Ken Banks who designed the software back in 2005 specifically for Non-Government Organisations working primarily in Africa.

Dr Fitzgerald is working on a new project with his colleagues to expand the farmer SMS network to include more production and marketing information.

(Further details on the project, including future plans for expansion, are available via a guest blog post from Robert here. The official University of Canberra press release is available here)

Mobile meets plough in Cambodia

In this, the tenth in our series of FrontlineSMS guest posts, Dr. Robert Fitzgerald – Associate Dean of Research at the University of Canberra in Australia – explains their use of the software to provide agricultural data to farmers in Cambodia, and their plans to expand the service over the coming months


“For the last three years I have been working with my colleague, John Spriggs, on agricultural development projects in Cambodia. John is an agricultural economist who has over the last few years been applying participatory action research methods to improving agricultural marketing systems in developing countries (Cambodia and Papua New Guinea). My background is in technology and education with an interest in the application of robust technologies to help users communicate and work together.

Originally, John invited me to join him on his project to explore ways we could use communications technologies to the improve the marketing system for maize and soy bean farmers in western Cambodia. From our early workshops in Battambang, Cambodia, we identified poor communications as one of the key constraints to improving the marketing system. At that time I had been developing an interest in mobile communications, particularly the application of SMS, and was following closely the work of the Pinoy Internet Farmers project.

Early workshop in 2006 with traders and farmers, Battambang, Cambodia (photo: Rob Fitzgerald)

Following an initial workshop we presented a proposal to the Australian Centre for International Agricultural Research (ACIAR) for a year’s extension to John’s project to develop the Electronic Marketing Communication System (EMCS). During that project we used a propriety system, Infotxt (developed in the Philippines) to run an SMS server from the Price Division of the Ministry of Commerce (MOC) in Phnom Penh. A price officer collected price data each week and used this to update the server.

The trial was very successful and the information on demand system allowed a user to text specific keywords and that would return price information. We conducted a number of training sessions in 2007 and later that year the Cambodian People’s Party endorsed the EMCS as a significant agricultural innovation in Cambodia.

During 2007 we worked closely with Pieter Ypma from CAMIS and for a number of months they used our EMCS system to SMS-enable their web-based price database. In early 2008 our project concluded and we handed the server across to the MOC.

A new project begins

In April 2008 we started a new project which focused on production and marketing problems faced by poor smallholder farmers in northwestern Cambodia. While still a comparatively fertile area, over the last ten years crop yields have been declining and soils are being degraded by excessive cultivation and burning. Much of the crop development has been largely driven by market demand in Thailand, however local farmers are disadvantaged by lack of market information, and inadequate post-harvest technology and transport infrastructure.

The overarching aim of the project is to improve the functioning of the production–marketing system for maize and soybean in north-western Cambodia as a key to increasing cash income, sustainable growth and poverty reduction for smallholder farmers. We have established eight village clusters, four in the district of Samlaut (a protected area containing the last remnants of tropical rainforest in Cambodia) and four in the municipality of Pailin (a previously heavily forested area known for its gem mining).

Local farmers at recent field day in Pailin (Photo: Rob Fitzgerald)

Farmer workshops have been investigating key socio-economic issues related to adoption of the improved crop technologies and improving land use. Village level workshops have worked on gross margin and partial budgets to examine return on investments for various production technologies and discussed access to marketing information.


Building on the previous project we wanted to extend our SMS work to both the production and marketing system. After a number of months of evaluation we selected FrontlineSMS as the platform for our SMS work (previously covered on kiwanja’s blog here).

In February 2009 we set up two FrontlineSMS servers. The first was with the newly formed Northwest Agricultural Marketing Association (NAMA) in Pailin which had particular focus on the provision of information (rated top priority by members) and the exchange and sharing of silo association price and market information. A second server was installed with the Maddox Jolie Pitt Foundation (MJP) in Battambang with particular emphasis on basic market information and health alerts. Check out the Pitt-Jolie press release here.

In June 2009 we further refined the systems and NAMA is focusing on using FrontlineSMS to broadcast messages to members, to provide maize, soy bean and cassava price information, and to distribute seed and fertiliser (input) costs.

NAMA FrontlineSMS server: Robert Fitzgerald with Moul Sam who runs the server (Photo: Rob Fitzgerald)

MJP has developed three main knowledge areas/teams for its FrontlineSMS server:

  • Health – Health alerts and health-related information, and health communications (e.g. clinic reminders for pregnant women)
  • Farmer – Price information, buyer contacts, weather information and farmer communications (e.g. meeting reminders)
  • MJP Head Office Field Communication System – A field communication system to provide field worker contact and updates, emergency/disaster protocols via SMS, meeting reminders, etc.

Future SMS applications

While our SMS servers are still in the early days of use we are exploring other SMS-based information systems such as GeoChat from InSTEDD and mhits– an innovative SMS-based micro-payment system developed in Australia by Harold Dimpel.

Volunteer and Intern opportunities

In recent conversations with both MJP and InSTEDD we have been exploring the possibility of establishing intern programs to encourage volunteers work with us. Some good work is already underway by volunteers. Oum Vatharith from Phnom Penh is already working on a Khmer translation of FrontlineSMS and I have talked to him recently about the development of FrontlineSMS user manual in Khmer.

We will be looking for volunteers in the form of software developers and IT savvy community/education development folk who are interested in working with these leading NGOs to help ensure that we realise the potential of technology to make a difference to peoples lives. If you are interested in helping out, please leave a comment!

Thank you.”

Rob Fitzgerald
Associate Dean of Research, Faculty of Education
University of Canberra