From the development team, I’ve already spied a few things that were being done by the field teams that go outside the standard deployment protocol (for example, calling up the LMMS mobile client application on Firefox’s platform is happening, I suspect, to circumvent the inherent speed challenges in the registration application as a result of having to rely on the Window’s Mobile operating Systems used by Intermec’s CN50. The ingenious nature of users to circumvent a slower process than they would like, is well noted … but just let me say that speed-penalty of Window’s Mobile it is in part a major reason for the delivery of the application on the Android platform which will be officially released for field use at the end of September (again – thanks in part to CIDA for helping us to devote resources for the build).

I do appreciate the feedback from users like Jermaine Bayas (Building Resilience Officer at Oxfam) who summarizes three key benefits that he sees of LMMS as:

1. Better data management supporting an aid distribution process with technology.
2. Helping Oxfam become more efficient.
3. Making the process of humanitarian action easier for the beneficiaries.

Well said Jermaine!

The last posting I put up, presented some metrics on the overwhelming support for LMMS expressed by  individuals from a broad cross-section of humanitarian organizations. As a result of that enthusiasm, World Vision submitted a funding request to CIDA late in the year aimed at encouraging agency trials of the existing LMMS system to help these organizations develop the critical knowledge and staff capacity to make effective use of this award winning innovation. It is my delight to announce that the proposal was successful and we have begun work in earnest with a number of agencies to meet three broad objectives. Namely:

• Extend knowledge and capacity to a wide number of agencies: Support the acquisition of hardware, offer the LMMS software as-is, help identify new skills that are needed by aid workers, train staff, educate a wide number of stakeholders in each agency as required, make available resources and learning from World Vision etc. etc., all in an effort to expand the participating agencies’ own institutional capacity development. The intent being to enable different agencies to have the skills and expertise to deploy the systems against what they have been designed for, such that each agency can take advantage of more efficient and accountable field practices.
• Grow the LMMS system itself against an existing backlog of features that have been identified and are known to be important for better humanitarian action. To encourage lessons and feedback from partner agencies on what works and what doesn’t. The latter will help strengthen the innovation where the demands are within scope and are technically feasible. To develop the system such that proprietary licensing from the private sector will not impede the deployments of systems like LMMS among the aid sector.
• Encourage research and evaluation of technology in the humanitarian space, studying the experiences across multiple agencies and with staff from different domains in how such systems can offer benefit. This work aims to assess the suitability of these technologies for humanitarian action and its impact on effectiveness and accountability.

I’ll be updating the progress, challenges and learning we take from these effort in the months to come. For now we would like to extend our appreciation to CIDA for taking the bold step  to support the principles of learning and sharing knowledge among agencies.

Its our delight to work closely with a number of agencies that already do amazing work in our combined goal of offering better humanitarian programming.  We look forward to learning from and contributing to the efforts of Medair and Oxfam GB who are now active partners in this exciting CIDA funded initiative! We expect more agencies to join the ranks of being LMMS’ers! Stay tuned and visit this page for the list of participating agencies.

And supported by:

At the conclusion of the session, we asked all participants from these 12 agencies, a few simple questions to ascertain the perceived value of the systems we had developed.

Q1: Did they see value for the humanitarian sector in the mobile technology software that we had developed?

Consolidated Responses:

Q2: Would they see immediate application of the software system we developed within their own operations (i.e. had we modeled the software in a way that had application for different humanitarian actors)?

Consolidated Responses:

Q3: Should we stop developing and growing the system or do we forge ahead with new features and functionality that are within the scope of what LMMS is designed for?

Consolidated Responses:

The last question was particularly interesting, because it also led so many of the trainees  (even the few that weren’t coming with a technical background) to ask about deploying the systems on consumer-grade devices over our existing commercial grade hardware devices (as you’ll see later, this did not have to be an either-or scenario). Moreover, those that had technical backgrounds were vocal in identifying the Android smartphones as the preferred platform of choice.

In our diverse group of aid workers, here was living proof that technology matters. Many of the trainees may not have been able to articulate why Android over other systems or to fully explain the pros and cons of consumer versus commercial grade devices. Yet all of them knew that mobile was important and could be applied inside the humanitarian domain. So why would (or should) humanitarian aid workers care about mobile technology?

Why Care?

I think the answer, lies in the fact that these people had correctly put their fingers on the pulse of the new world reality for humanitarianism. Namely a world in which the mobile technologies are presenting a paradigm shift in how we work, a world that is becoming increasingly networked, and a world in which the potential to access lots of data offers tantalizing opportunities (as well as a zillion risks – but that’s another blog entry in itself).

Let me try to expand on some of these dynamics with facts that you may find interesting and that may explain why these participants cared about mobile and perhaps why you as an aid worker, should. This will set the scene for a major announcement at the end of this blog entry about LMMS’s mobile capabilities!

We’ll start with analysis from Morgan Stanley based on data of computing cycle usage patterns from the era of old mainframes to the 2010 reality. This analysis shows that each time we experience a major computing shift, we see essentially experience a 10 fold increase in the number of devices that are connected to the Network.

Consequently in 2010 they predicted a massive growth of mobile “consumers” of devices and network systems as shown below:

Source: Wroblewski, L. (2011) presenting at Tech Master Class and based on Morgan Stanley Technology Research by Meeker, M., Devitt, S., & Wu, L. (2010). Internet Trends.

Morgan Stanley were essentially projecting more numbers of people on mobile devices than there were people on earth. How would that be possible? The answer they reasoned lay in consumers like me who own multiple mobile devices of course. The same research predicted that by 2012 the total global shipment of smartphones would surpass that of Desktops, PCs, and Notebook PCs.

You may ask how these predictions fared?

• In Q4, 2010 Smartphones passed PC global shipments for the first time in history. That was 2 years ahead of Morgan Stanley’s predictions!
• Just last month IDC reported a 13% decline annually in the first quarter for PCs, while tablets were up a whooping 142% and smartphones up at 41.6% indicating a continuing radical shift to mobile.
• ITU reported that mobile subscriptions were already at 6 billion by 2011.
• In 105 countries, there are now mobile subscribers than there are people.

Consider the Philippines for example. In 2012, the mobile phone penetration rate stood at 103 per cent – that’s more phones than there were Filipinos. In Pakistan (the second largest recipient of international aid assistance between 2006 – 2010), 68.8% of the population are subscribers to mobiles.

In 1993, the number of mobile subscribers in Africa was in the order of hundred of thousands. In 2003, it was 4 million. And in 2012 it stood at 735 million people. The BBC reports a 70% penetration rate in Africa for mobiles. And the ITU (Ibid.) published data showing a 79% penetration rate of mobile in the developing world. Evidence again that when it comes to mobile, the rates of change are astronomical.

In the developed world, the mobile paradigm shift has had massive impact on other sectors. Wroblewski (Ibid.) for example, presented data from segments of the economy including that of e-commerce, wherein Amazon recorded USD 1 billion being spent solely via mobile devices. In the Social Media segment, Facebook recently announced first quarter 2013 revenue earnings for the company in which they described a 23% increase in mobile active users (year-over-year)  resulting in significant mobile advertising revenue for the company and silencing many naysayers of FB who questioned the company’s ability to operate as profit generating business.

The Connection to the Humanitarian Sector

More mobiles devices in all corners of the world means what exactly for the humanitarian sector?

As I’ve previously stated, I don’t believe that technology will eradicate the underlying causes of suffering and poverty. But it is a powerful enabler to helping in the cause. Mobile gives us an opportunity to influence operational needs like never before.  For our purposes the shift to mobile technologies puts new tools in the hands of beneficiaries and aid workers. This has had massive ramifications in terms of communication capabilities and access to information (think of the impromptu efforts to use social media and other software platforms to support crowd sourcing and crowd seeding).

There are also host of data collection tools that allow for simple data extraction by aid agencies. Other uses from mobile data had lead to really innovative mechanisms for humanitarian needs (such as tracking population movements using phone locations in response to diseases like Cholera in Haiti, which was subsequently shown to be an effective early proxy-indicator to the epidemic).

Further, as in the case of LMMS, it also means empowering aid workers with mobile tools to access the software we developed that models real aid-worker operational practices (which then results in a far better humanitarian work on the ground, which then has a host of  positive trickle-down effects for aid recipients).

When we first started with LMMS, we could not rely on connectedness of networks. But today, we are seeing those barriers starting to come down. The connectedness of the world is growing leaps and bounds as shown in the diagram below:

So – What Do You Make of All This?

Let’s distill this down to a simple picture:

1) Mobile technologies are part of the de-facto way in which we now live and work. If “mobile” is not in your nomenclature, you will be missing out.

2) These technologies are becoming cheaper over time (and the cost of embedding Internet capabilities in many emerging technologies continues to fall  – in a few years, we’ll be seeing wearable computing devices that go far beyond our current concepts of mobile). This means even more application potential within the developing world and inside the humanitarian domain.

3) The average smart phone today has more computing power in it than NASA used to send Neil Armstrong and the Apollo 11 crew to the moon. The capability to leverage such power in portable devices means the ability to radically alter how we work.

4) Connectedness is making mobile web computing far more feasible in every corner of the world (in part this explains the architectural evolution for LMMS – pleases tuned for announcements early next year on this front!).

Raise the Curtain: LMMS on ‘Roids!

The simplified picture painted above in conjunction with what aid-workers actually want (including those trainees I started this blog entry with), led my development team to put in an enormous amount of work over the last 8 months into our systems. This has culminated in a version of LMMS that now supports the Android platform for both consumer grade devices (smartphones) and soon-to-be issued commercial grade devices running Android 4.0+.

Moreover, in the development we have been doing, we have also committed to  support the existing Windows platform running on the current commercial devices for the next 12 – 18 months.

This means that the super feature enhancements to be found in the LMMS mobile applications can be enjoyed on both the existing Windows Mobile environment and the Android operating system thanks to the hard work of the development team.

The design phase enabled us to also factor in feedback from expert humanitarians (on things that made sense and that were technically plausible!).

The results show in such things as a far cleaner beneficiary registration application with capabilities to invoke GPS data collection at locations where people may be. The application tracks migrant information (where people came from in cases of displacement). In addition there are a slew of revamped vulnerability tracking capabilities. The latter includes new household level and member level vulnerability conditions and have resulted from our extensive analysis on what vulnerabilities actual are and how they relate to people and families. This analysis looked at traditionally tracked vulnerabilities inside the aid sector, many of which were ambiguous, repetitive, difficult to differentiate between household and individual conditions, and; according to field users, just plain confusing. In the full release of the next major upgrade to LMMS (due later this summer), these vulnerabilities will show the amazing advantages when using mobile technologies in the field – including machine logic to derive certain sets of vulnerabilities (thereby enabling more accurate and even faster data collection all of which can be reported against)!

The move to the Android system has also thankfully mitigated the slow performance of Window’s Internet Explorer that had my team pulling their hair out over the last little while (FYI Explorer was the only viable option on the Windows Mobile platform running on commercial grade devices). Faster processors in the smartphones and just the nature of Android (being Java based technology onto top of a Linux kernel) enables blistering speed in the registration application today; constrained only by the speed of the user’s Internet connection.

By working on the Android platform, LMMS now offers a full application feel that includes tight integration with the device’s camera and scanning capabilities in an embedded and specially designed Android application developed by the LMMS development team and codenamed LAB.

The upcoming NOMAD event will demonstrate a version of the Android system to attendees.

The full version (inclusive of the aforementioned vulnerability additions) will be released later this summer and I’ll post details on this site when that happens.

So – as an aid-worker, I implore you to embrace your smartphone (but remember to put it down when you are having dinner with your spouse – that’s just rude). Come talk to us about getting LAB and its functionality available inside your agencies. If you are in Paris May 15 – 17th, please join my colleague Richard at NOMAD for a first hand review of the Android system. Members of the World Vision community will see LMMS on ‘Roids at the Global Relief Forum (May 8th) after having given Otto and Keith Chibafa a crash course in the systems.

Enjoy and peace.

If you are in Paris May 15 – 17th, why not join many others who will be heading to NOMAD’s workshop aimed at linking solution providers with humanitarian organizations?

While there be sure to stop off and visit the LMMS representative. We plan to demo some exciting new product developments that followed directly from the feedback we received from INGOs, UN and intergovernmental agencies that underwent a half-day training with the LMMS tool. While at demo you’ll also see how LMMS goes beyond data collection and moves to empowering automation (and better humanitarian practices) directly in the field where aid workers are active.

Stay tuned and I’ll make some more announcements pertaining to the unveiling of fab product features closer to the NOMAD 2013 workshop date. Registration for the event can be found here.

Data from LMMS systems can also be used to better study and communicate our actions to a wider stakeholder group.

As an example I thought I’d present some information on World Vision projects being run in Laisamis, Kenya where LMMS has been used to manage a number aid and beneficiary programs. This is a region; in the center of the Kenya as shown by the LMMS mapping function below, has been severely affected by drought in recent times.

The Canadian government (through CIDA) and World Vision have helped to implement a variety of health related interventions. The local field teams at these last mile locations have been using the LMMS data for their day-to-day activities and also collecting some useful health metrics. Data from three roaming servers was polled (a snapshot data taken at the beginning of this year) was looked at to help describe the humanitarian interventions and also to demonstrate the potential in generating real data for evidence-based analysis.

Here’s some useful info on the information collected in Laisamis:

• Server 1 held data from 1,704 households with registered family members of 3,340 people, of which 2,508 were females and 832 where males.
• There were 879 instances of health metrics collected from the the “last mile” collected between Feb 2012 and November 2012.
• 15 Projects were run and 5 distributions that included the distribution of water purification agents.
• 145 people were found to be sufficiently vulnerable to be identified as primary beneficiaries.

• Server 2:  230 households with registered members of 475 people, of which 199 were female and 276 were male.

• Server 3: 3,695 households with registered family members of 6,732 people, of which 5,144 were females and 1,588 males.

Looking at some visualizations of the data .. here’s an example of the vulnerability plots based on the data sets of the three servers. Note that things like the displacement status will be zero as this is not an intervention that involves migrants. Also note that the vulnerabilities listed are based on optional data that can be collected by the LMMS system but are not enforced by the systems (such that field monitors can collect if its deemed important, and of course once people are registered, then updating such information is a breeze):

For a breakdown of the number of under 5 year children that were registered in Laisamis across the three servers and their breakdown by gender, here’s a handy viz:

Digging a bit deeper into the available data
Using open-source statistical packages, you can start to get some idea of the data analysis potential (please note that I have not cleaned the data entry points and I’m just presenting this as information of the potential to help visualize and stimulate research-based evidenced coupled with actual last mile data).

Looking at the data collected under Server 1 from a certain region of Laisamis, we can generate some neat data visualizations to describe over 800 health monitoring data collection points based on a mother-and-child subsection of people registered using LMMS.

A box plot on age profile shows that the bulk of these health interventions at this “Last Mile” was with young children under the age of 5, and some with vulnerable mothers (look at the above presentation on vulnerabilities):

The   descriptive stats on the data points collected is shown below:

Min. age as 3 months

1st Quartile as:  1.000  years

Median Age:  2.000   years

Mean Age: 3.514  years

3rd Quartile:  3.000  years

Max. Age: 42.000 years

I ran a quick visual on Weight for Height Scores modeled on time to see if there was “indication” of change (i.e. hopefully improvement on z-scores):

Looks as if over time, there has been some improvement in the collected WFH metrics, but as to whether that is statistically significant or not, would require more digging. For example, are the observed “improvements” over time in reported WFH Z-scores due to natural weight gain expected from maturity or are they linked to the extra caloric rations being provided? (Again – please note that I didn’t clean the data and simply ran the stat visualizations as examples of what can be looked at in the data).

Rather than looking at the WHF measures, I also ran a faceted density plot to visualize the different spreads of data points of MUAC (Middle Upper Arm Circumferences) measures, grouped by the categories of Gender is shown below:

 Is there difference in the measures between males and females, who are mostly under 5? Well perhaps there is difference in the reported density plots, but again – the data would need to be reviewed and a statistical interference on the notion of significance.

As noted by the UN’s World Food Programme’s Beatrice Tapawan:

“A first in Southeast Asia: Pioneering ‘Last Mile Mobile Solutions System’ to enhance accountability to Typhoon Bopha affected persons

The Food Security and Agriculture Cluster is leveraging innovative technology to strengthen humanitarian assistance in Typhoon Bopha affected areas. Early recovery initiatives are piloting the use of ‘Last Mile Mobile Solutions System’ (LMMS), a digital tool that aims to enhance transparency, accountability, and efficiency in delivering assistance. The introduction of this wireless technology is a first not only in the country but also in Southeast Asia. An innovation of World Vision, LMMS consists of a handheld device and a laptop and can be used virtually anywhere. In the hardest hit areas of Davao Oriental, Compostela Valley, and Agusan del Sur, initial signs are already promising.

Traditionally, food-for-assets and cash-for-assets programmes entail a meticulous paper-and-pencil process to collect vital information, such as participant names, vulnerability categories, and entitlements. The automated system essentially streamlines and digitizes these data. Then, participants are provided with a photo ID card with a barcode. During the ‘last mile’ of the process wherein participants claim their entitlements, all it takes is a quick scan of the ID card and upon validation, food or cash is transferred immediately. Confusion is minimized while security is enhanced. The right entitlements are received more quickly and by the worst-affected people. A sense of dignity is preserved among beneficiaries who spend a significantly less amount of time waiting in line. In addition, the system allows for rapid generation and sharing of disaggregated data and participant lists to improve basic services and to minimize duplication of efforts among the larger humanitarian community.”

From our partners in the field here are some pictures of the actual deployments:

Sorry – no time for audio .. you’ll have to invite one of us to talk you through the prez!

Place the mouse over the visualization to get even more embedded details. And please let me know what you think of this material!

Combining software applications with a hand-held device, LMMS simplifies remote data collection, beneficiary management, commodity distribution and reporting, with a particular focus on food and cash programming. Staff track distributions using electronic records, rather than with pen, paper and ink thumbprints of recipients.

World Vision is using LMMS in the community of Corail, where Ms Belance was helped to resettle in April 2010 after the devastating Haiti earthquake of 12 January 2010. World Vision has also implemented the system in other key intervention locations, both in Haiti and in other countries.

“Things go so fast with this card,” she says. She recalls her hours of “calamity” back in the beginning of the response. “Before this card, I would wait up to three hours in the sun to be served.” Ms Belance has been receiving multiple types of support through World Vision’s efforts to address the needs of displaced people living in Corail.

On top of making it easier to serve beneficiaries, the LMMS system gives a sense of dignity to Haitian cardholders who lost identity papers during the quake in a country where citizens’ identification was an issue even before 2010. “When the paper-based system was replaced by this digitised one, I was relieved,” says 22-year-old Isaac Cardichon. He no longer has to argue to prove that he is the actual beneficiary on a list.

“With LMMS, no one else can get what is dedicated to me,” echoes 29-year-old Wydelande Saul. Compared to the beginning of the response when beneficiaries were required to place their thumbprints on papers during distributions to receive what they were owed for food- and cash-for-work activities, LMMS streamlines the process.

“This is how an NGO treats people with dignity,” affirms Pierre Poniathiosky Sanon, a beneficiary who headed one of the four blocks in Corail. He remembers that early distributions often turned into fights between camp residents. “With the old system, many would try to benefit more than one time,” he says. “And that’s how a conflict might occur.”

“It’s not just about preventing fraud; it facilitates our work too,” states Yousely Jacques Simon,  information technology officer, LMMS – WV Haiti, while mentioning that reports can now be completed and shared just minutes after a distribution using a hand-held computer.

By streamlining remote data collection, beneficiary management, commodity distribution and reporting, LMMS ensures that humanitarian staff at the “last mile”—where aid reaches people in need—can focus on programme quality rather than cumbersome paperwork, which is more vulnerable to human error and corruption. In Haiti, the LMMS system has helped to more accurately track food rations and distributions of cash and supplies; it has also enabled teams to generate reports immediately for donors and other key stakeholders.

“Working in the difficult context of Haiti, such an achievement could be considered unthinkable; but that is exactly what the LMMS system is designed for,” says Keith Chibafa, LMMS manager – WV Haiti. Mr Chifaba runs a unit of seven staff which oversee and facilitate the use of LMMS, which was once a pilot project but is now a standard tool.

“The LMMS system is a game-changer,” Mr Chifaba affirms, “At all steps including planning, registration and reporting, things get done faster and more accurately.”

Written by Jean-Wickens Merone, communications officer – WV Haiti