iSTEP Tanzania

Author Archive

Today is the official last day of the iSTEP internship. While the team in Tanzania won’t be hitting home soil until Tuesday, our formal involvement in the projects ends when we pack it up today.

It’s been a fun 10 weeks, a hard 10 weeks, an exhausting 10 weeks, but it certainly hasn’t felt like 10 weeks. The iSTEP team has braved monkeys, lions, snakes, mosquitoes, tornadoes, and hair cuts. We surprised no one more than ourselves in the progress we were able to make across the three projects. We’ve learned more than we thought possible and made countless memories.

One might think I probably had the boring job in the team being home in Pittsburgh but I had my share of excitement. I’ve worked myself to exhaustion on more than a couple days and I’m proud of the work the team has done and hope to continue my own work in ICTD in the future. The ever-changing day to day conditions, the wealth of viewpoints, and the projects I got to work on more than kept me interested. I hope the things we’ve started this summer continue to develop and make a difference in the lives of Tanzania.

And so it is without any eager that I leave my humble desk that was home to me over the summer. I would say that I don’t know how I will manage to keep myself busy but then I remember classes start at Carnegie Mellon in a matter of weeks. But then, homework just won’t be the same.


Everybody on the team is quickly getting short on sleep. I have found myself working in the AM on more than one occasion and I have caught a couple of the team in Tanzania in what should be early morning time as well. There’s only one week left and everyone is clearly feeling it.

So what’s left to accomplish? We’ve discussed on the blog a lot of work that’s been done and I hope by now you’ve got a good impression on the three projects. A lot of the late nights have involved putting in as much extra functionality as possible so that when we leave it’ll be in a state such that people on the ground will be able to utilize the technology that we’ve been developing for some time. While we’ve spent a lot of time developing recently, slowly we’re starting to shift into a more presentation oriented outlook. Everybody is discussing and developing reports and slides so that we can present to the UCC and students back home in order to show off the work we did and how it’s useful. It’s a nice time to reflect on where we’ve been.

In developing my portions I was able to go back and look at the first reports we did before the team left. It was interesting to revisit the places where we started. Granted a lot of that information continued to be relevant throughout the duration of the project but looking back and remembering some of the research we did, attempting to find relevant work from which we could draw inspiration. CMU HealthLine, Project LISTEN, Project Kané, and countless others. From all of these projects we were able to draw inspiration and learn something new. It’s unlikely we’ll make as much of an impact as things such as Project LISTEN which for over a decade has been improving the way children learn to read, but we hope that when we leave we’ll be able to tell our experiences such that other people will be able to benefit as much as we have benefited from those before us.

Almost everybody has gotten to the point where their projects are demo-ready. Hatem performed the literacy tools demo today showing off the basic soccer game, Dan did a demo for the social worker application a couple days ago and Brad will be testing our new method of mode-switching on the Braille Tutor in the coming week. The program is nearing its final weeks. We’re far from done but at least everything has taken shape.

A big part of my week has been spent on graphic design for the literacy tools game. Just as a preface, I’m not a graphics person. There have been many things I’ve had to learn on the fly for this program but few have put me more outside of my comfort zone than this project as I am far from an artist.

Suffice it to say I have read just about every isometric pixel art tutorial on the internet. Pixel art has been gaining a following, you may have seen examples of it if you play Habbo Hotel or have visited It utilizes the fact that digital graphics are displayed as a collection of small colored squares to make up a large object. When you consider isometric pixel art specifically you consider pixel art when shown at an angle to further illustrate the illusion of depth which when mastered can create stuff like this. However given my week of practice I am not quite at that level yet.

The idea we’re working on for the literacy tools game takes advantage of the immense popularity of soccer – or football – to encourage students to practice their reading abilities. In so doing we have the player in a penalty kick scenario facing off against a goal keeper. The player is asked a question and chooses from a collection of prompts. If the player chooses correctly, they score a goal. An incorrect choice and the ball is either caught or flies wide of the goal.

With that altogether, a large part of my week produced this:


It’s not the prettiest but hopefully it’ll get the job done. If you are a pixel artist and have any tips or know someplace where we can get some, please help us out and leave a comment!

There’s only three weeks left in our internship program and I think everybody is feeling it. We’re heavily into the development stage right now and we’re working hard to get the projects up and running, as well as beginning the thought process of how we’re going to leave Tanzania at the end of the month.

The past few weeks we’ve been focused on development and this week was no exception. Hopefully, based on my and Hatem’s posts last week, you’ve gotten at least a slightly better understanding about the social workers and literacy tools project. I’m hoping here to summarize here a bit of our work in these past few weeks in adopting it for Tanzania.

The Automated Braille Writing Tutor is a TechBridgeWorld (TBW) project that has been developed over the past few years by various students, staff, and professors. You can visit TBW’s page about the braille tutor at The Braille Tutors that Brad brought to Tanzania are of the version 2 design, pictures of which you can see on the website. This design involves a large six-button cell for beginning users, and two rows of smaller cells for more advanced users. The tutor detects input from the learner and can give feedback with a variety of different modules, including a game that plays an animal sound and asks the user to spell the accompanying word.

Thanks to the work of Imran Fanaswala at Carnegie Mellon University in Qatar as well as other students, the Braille Tutor is growing to support a multitude of different libraries. When we received the tutor’s library, it already had support for Arabic and French Braille, and the modules were ready to be adapted for any modifications we made.

Therefore, we’ve been attempting to integrate Swahili into the tutor. We have translated much of the actual text for the games and learning modes (Did you know Simba is Swahili for lion? The protagonist in the Lion King is actually a lion named lion!) We’re also recording voices of native Swahili speakers to ensure that the audio playback as clear as possible.

Additionally, the growing functionality of the Braille Tutor has led to the development of a new but not entirely unwelcome problem. With six modes originally, the Braille Tutor that was taken to India last summer utilized the large buttons on the face of the tutor to switch modes. This was simple enough since the tutor had just six buttons and six modes. However, by the time the interns left Bangalore, they had designed two new modes. When support for utilizing multiple languages is introduced, one effectively multiplies the number of modes, which makes things even more confusing.

Brad has organized multiple meetings for the team as we brainstorm and try to conceive of the best way to make the tutor easy to use while continuing to increase its functionality. We have created two designs that he will be demonstrating in Tanzania soon, both of which have a scroll style menu design where a user steps through all of the different modes by pressing one button and then pressing a confirm button when reaching the desired mode. However, one of the designs will implement a hierarchy such that one can choose a language, a type of lesson, and then an individual lesson, so that the tutor is better prepared for the growing modules that continue to be added.

So by now you probably have a good idea of the three projects. Hopefully the people for whom we are developing this technology will have a good idea of this as well in a couple weeks.

This week seems to have gone by awfully quickly. That hasn’t stopped it from including a lot of work, but at least it’s Friday and the 4th of July is tomorrow. A celebratory weekend is something I could really use right now.

The bulk of my work this week has been trying to figure out the situation with the social workers application. The basic idea behind the application is based on the fact that social workers in rural areas of Tanzania have to submit forms to organizations at the national level about their work in the field. Given the resources available in these rural areas, submitting forms can be a time consuming task and can take over a year in some cases to make it end-to-end. Without Internet access, how can we expedite this process? In Tanzania, Dan has been heading up an idea with cell phones.

Cell phones are one of the more prevalent technologies that have emerged in Africa. Given the prevalence of cell phones, the first few weeks of this internship involved researching what the state of their prevalence was in Tanzania. We researched carriers, costs of different models, and how people used them. Eventually, we hypothesized that using SMS messages, rather than voice, to send form data is more cost-effective due to several factors. Moreover, the usage of SMS in Tanzania is very popular, so if the service we render is simple to use, we will be able to get useful data back.

This is not the first time that SMS messages are being used to relay data. In some parts of the world, one can handle financial transactions via SMS messages sent directly to one’s bank, while farmers are using text messages to track market data to know the best place to take their produce for sale. Knowing that other projects have been utilizing SMS as well with excellent results made us more confident that SMS would be an ideal avenue for transferring data.

That being said, our recent work has been about developing the back end, the software that will receive and track the data that gets uploaded by the people in the field. The people at Kannel have a back end that we plan to utilize. It allows us to receive text messages from a phone connected to a computer and access its content in a variety of formats. From this example, Dan has been plugging away with Zope, an open-source content management system that we can use to collect, process, analyze, report, and export the data, sent by the social workers in the field. Zope is a pretty robust suite of tools that has given us a lot of functionality that will be valuable for a final product. However, as Dan said to me a couple days ago, “you get what you pay for.”

That is, Zope is free and open source, and this has meant that this week we’ve had a serious problem getting all the parts to work together. Zope itself doesn’t come with a ton of functionality, but rather its usage is augmented by all of the modules that are available online. However, given the wealth of developers and lack of communication between people developing for Zope it’s not always true that any two particular modules will work together. Additionally, with the modules constantly being written and rewritten, two different versions of the same module can mean the difference between a functioning application and one that won’t start. Zope tries to help users out by listing some Known Good Sets of modules that should work together, but if one wants anything not listed or needs a newer version of a module than that which is listed in the known good set, he or she may encounter more than a few problems.

Downloading multiple versions of modules to try to get them to work together has been a painful process for me and even more difficult for those in Tanzania, where the Internet connection isn’t always stable. Luckily, by Wednesday, most of these errors were cleared up and we are now getting to more and more of the breadth of functionality that Zope provides. We hope to have a demo very soon that we can get feedback on.  Plus, if we utilize this solution, our software costs are zero.

It may have been a pain to get this far but it’s difficult to complain if we do indeed get results at no cost.

First of all, congratulations to Hasheem Thabeet who last night became the first Tanzanian player to be drafted into the NBA. After being born and raised in Dar es Salaam, Tanzania, the same city that our team is currently stationed, Thabeet attended the University of Connecticut before being selected with the second pick of the 2009 NBA Draft. Good luck in Memphis, Thabeet!

While last week we battled the storm to get back into the office, this week we encountered a different sort of storm. Looking back to Timi’s post last week you can see that the ground team has been meeting with a lot of the people who will eventually be using whatever we manage to develop. Through interviews with various parties involved we’re finally beginning to collect a comprehensive set of needs that we can try to address. What this means is that on Wednesday we examined these and put together an intense brainstorm session where we talked and threw around ideas about what and how we can try to solve. It was an invigorating session and when so much of our conversation has been over e-mail, to have everybody from our advisers in TechBridgeWorld to the members on the ground in one conversation produced a lot of interesting dialog that still continues to permeate e-mails days later. Every day since Wednesday I’ve been returning to my notes from the meeting and feeling a bit overwhelmed at how much came out of the meeting and trying to decide how much of it can actually be accomplished. Regardless, next week I believe we’ll be having another meeting and I hope it’ll be as fruitful as this one. If anything was clear from this week it’s that we’ve got a team with a lot of creative ideas, the next 4 weeks will show whether we can actually do something with them.

So as Emily mentioned, Pittsburgh got hit with a bit of a storm Wednesday night. I received the pleasure of finding a tree torn down and laying in the road on my way home from dinner and Thursday morning learning that the key fob for my building was malfunctioning so I had no means of entering the building should I wish to leave. Further Thursday morning we were informed as Emily mentioned that Newell-Simon was flooding, we happen to be working in the lab on A level so there was a bit of fear about the equipment we’ve left in the lab but were later informed that only the power was out.

While last week I learned how to adapt when giving a presentation, Thursday I learned how to adapt when your office is being torn apart and you have deadlines. We don’t have the luxury of being able to delay progress on the projects as we only have 10 weeks to accomplish our goals and the team in Tanzania is not restricted by such stipulations. It did limit the amount of work I was able to do however as a lot of my equipment was still in the office. I also tried to access CMU’s library so that I could attempt to research ICTD projects to get some sort of perspective on other projects that have been tried in these areas but the VPN was also down. Needless to say, Thursday was a little bit frustrating as the papers accessible by Google are much smaller than the papers accessible by CMU.

Friday we were let back into Newell-Simon and while walking down the hall I found that the sink in the kitchen had begun flooding black water all over the floor.

It’s been an adventure.

Luckily this doesn’t stop our progress today. I was able to retrieve my Braille Tutor and save it from whatever further mayhem this weekend may bring and have been isolating the areas to change when we translate it to Swahili. The biggest part has been preparing for the project to streamline the input of forms for social workers in Tanzania. I’ve been working with Dan to collect a set of tools to consider for the project. We’ve looked at Kannel, EpiHandy, EpiSurveyor, FrontlineSMS, amongst countless others. Luckily I think we’re getting close to demo-ready for the project which will go a long ways to see just what is workable in the long term. The next few weeks are going to be a wealth of getting feedback and reacting. It’s scary but also exciting as I’m not sure how my days could get more dynamic at this point.

Twitter Feed

Error: Twitter did not respond. Please wait a few minutes and refresh this page.

February 2017
« Aug    
The opinions expressed by the bloggers are theirs alone, and do not reflect the opinions of Carnegie Mellon University, TechBridgeWorld or the iSTEP program, or any employee thereof. Carnegie Mellon University is not responsible for the accuracy of the information supplied by the bloggers.