Robots Podcast #191: TechBridgeWorld, with M. Bernardine Dias
In this episode, Audrow Nash interviews M. Bernardine Dias, Associate Research Professor at the Robotics Institute at Carnegie Mellon University, about TechBridgeWorld. TechBridgeWorld in an organization, founded by Dias, that develops technology to help serve developing communities. This interview focuses on a device that helps the blind learn to write.
TechBridgeWorld would like to acknowledge the Qatar National Research Fund (NPRP #30-6-7-91), the Fetzer Institute, and individual donors for making this work possible.
M. Bernardine Dias
M. Bernardine Dias is an Associate Research Professor in the Robotics Institute at Carnegie Mellon University’s School of Computer Science. She is also the Founder and Director of the TechBridgeWorld research group at CMU. Her research expertise are in technology for developing communities, assistive technology, and autonomous team coordination. Dr. Dias is a native of Sri Lanka. Her career goal is innovating means of developing and disseminating suitable and sustainable technology for empowering underserved communities. She is a strong supporter and mentor for women in science and technology.
- Download mp3 (18.1 MB)
- Subscribe to Robots using iTunes
- Subscribe to Robots using RSS
- TechBridgeWorld website
- Braille Tutor project website
- Braille Tutor SourceForge page
Audrow Nash: Welcome to the Robots Podcast. Can you introduce yourself?
Bernadine Dias: I am Bernadine Dias, and I am an associate research professor at Carnegie Melon University in the robotics institute. I run a research lab, which I founded there, called TechBridgeWorld.
Audrow Nash: Can you tell me the goal and motivation around your work?
Bernadine Dias: Our goal is to essentially re-think what technology can do for the communities around the world that aren’t currently served by technology – primarily communities in the developing world and communities with disabilities.
I grew up in Sri Lanka. I was born and raised there in a low-to-middle income family and was very lucky to get the chance to fly to the US and get an education in technology. Everything I ever learned about technology was at college. It’s been an interesting trajectory for me, and I’ve always been interested in understanding technology and becoming an expert in it so I can actually change technology to fit communities rather than forcing communities to change to fit technology. That’s basically our strategy and our motivation.
There are a lot of exciting things happening in the field of robotics, as I’m sure you would agree, but most of what was happening about ten years ago when we started this endeavor was only benefiting about two billion people (if we’re being generous). That’s not even half the global population. In our lab we’re trying to understand how we can change technology in ways that are accessible, relevant and affordable to the rest of the world.
Audrow Nash: Can you give an example of these kinds of projects? How are you fitting technology to people?
Bernadine Dias: We start by talking to people about their community, and about their vision: what their hopes are, what their dreams are and what is preventing them from getting there. We don’t ever want to go into community and tell them how they should live their lives. What we bring is our technology expertise; we rely on them for everything else.
For example, in 2006, we were introduced through a series of friends and connections to a small school for blind children in India. This school is pretty amazing. It’s a private school run by a woman named Miss Muktah, who herself had a disability due to an injury. Through her rehabilitation she met a lot of people who are blind and realized how difficult their lives are, so she started a school in order to help kids who are blind have a better education and access to a better life.
She raises funds through donations and then goes out to poor areas around Bangalore and slums where kids wouldn’t really be able to afford any form of education. The school is residential, and she brings them in, of course with their parents’ consent. They live there and they learn everything: life skills, health, hygiene, navigation, and she gets them through as far as they can get in their education system. It’s really amazing. A lot of the teachers there are also blind.
We started talking to them about what they do and we were very impressed about all the things they’ve accomplished. We were looking for ways we might be able to play a role. Those initial conversations were really to tell us more about the school, and then they got a local volunteer to get video of the classes – because this is all remote – and they would send that back to us.
One of the key things we are identified was that the students struggle to learn how to write braille, and braille is the only way blind people can be truly literate. In the developing world it’s especially important that they’re literate because they don’t have access to braille books, or audio books for that matter, or anything in terms of educational resources. How it works is that you learn to write braille in first grade, and then as you progress, the teacher tells you things and you take down notes in braille. Then you study your notes and you take your exams, and that’s how you pass out of the different grades in school. If you don’t learn how to read and write braille then you don’t get an education. It’s a pretty huge issue.
To make things worse, in the developing world students write braille using a slate and stylus. Braille is basically a tactile way of reading: the configuration of embossed dots on the paper maps onto either a character, or a contraction of several characters, depending on which grade of braille you’re writing or reading. To write braille, you insert a sheet of paper into the plastic pocket of the stylus, and you push down on the dots of the stylus to emboss the paper, and then you slide the paper out, flip it over and read it.
Now, because you have to flip the paper over, you have to write the mirror image of what you read … so if you read from left to right then you’re writing from right to left. Essentially you’re learning three alphabets at once: you’re learning English (Kannada is what they learn in this school, but they also learn English or Hindi or Arabic or whatever); you’re learning the dot pattern of the read version, which is typically a direct character map translation from each character; and then you’re learning the mirror image of that for writing. For a little kid in first grade this is a pretty stunning endeavor to undertake, and they don’t really understand why it’s important, so a lot of them to end up being literate.
We looked at this problem and we realized there were a lot of things we could do with technology, but we had to think very carefully about what exactly would be a tool that would be helpful.
One option would have been to eliminate braille all together … we could have come up with some form of audio system so they can avoid needing to learn to write braille. But this is not sustainable; when the kids leave the school they would not be able to afford something like this for a long time – maybe decades.
Another option would have been to create a tool that lets you eliminate that mirror image. In the developed world there are already digital braillers that let you write the same as you read, and it does the flip for you, and there are six key braille typewriters that you can type with. But even this technology can’t compete with the price point of a plastic slate.
We decided that the most appropriate and sustainable solution was to build an electronic tutor that would help make learning to writing braille using the slate and stylus more fun and effective. We call it our Braille Writing Tutor. We’re on our fourth version, and we’ve been developing it with this school and also other schools around the world.
It’s essentially an electronic slate that speaks to the child and gives them feedback while they are writing braille. We’ve incorporated games and curriculum to make it more fun and interesting.
Audrow Nash: What does the device look like?
Bernadine Dias: There are a few different versions. The version for young children who haven’t got fine motor skills looks like a box with six big buttons. They start by learning the configuration of the braille cells, the dot numbering, and the patterns.
The next level up is still a box with buttons, but it has three braille cells, which are a little smaller now, and underneath that there are two rows of holes for the six-dot system; you press the stylus into the holes to connect the circuit. You’re trying to scaffold their learning and their motor skills.
The advanced version is essentially like a slate. It’s a box with a lot of holes in the six-dot configuration, and the goal is to get the kids to just use the slate and the stylus, so that eventually they can move onto the actual slate.
Audrow Nash: Can you talk a bit about the development process of working with the community to create these devices?
Bernadine Dias: Our methodology is to spend whatever time it takes to really understand each other. It’s not just about learning from them, but they learn from us as well … who we are and our process … and that takes time.
When we come to a point where we think the parties involved are in agreement on what is worth trying, we do a paper prototype or a sketch, depending on who the community is. If they’re literate we can write; if they’re not then we try to do things diagrammatically. If they’re blind, sometimes there is braille involved, or we will build something physical with paper or cardboard that they can feel.
We start with this initial prototype stage so that we can get their feedback early. This is really important because if you give them something too advanced, in some cultures they won’t say anything negative about it; they feel like you already did this pretty amazing thing and they don’t want to be negative … they want to be polite. People are really nice.
Audrow Nash: They won’t give you honest feedback on the device … ?
Bernadine Dias: Exactly. The more rugged or ‘prototypish’ it looks, the more they’re willing to say “Oh wouldn’t be nice to add this?” or “This part probably won’t work so well.” That whole conversation – that iterative process of designing with the community – is also good for us because it ends up building more trust, and it gives that joint ownership of the project.
When we feel that we are at a point where it’s worth going for a more complete prototype – it could be software, it could be hardware, it could be both – then we run field tests and iterate again. In some places that can happen really fast, in others it takes years. You just have to go with the flow of what is best for that community, because you’re taking people’s time and energy (including yours), and it needs to fit with everybody’s lifestyle and whatever they’re investing in the project.
Audrow Nash: You mentioned that when you were working with the school in India, you were often remote. What challenges were there?
Bernadine Dias: Oh gosh, everything! The school in India is actually one of our easier communities to work with because, even though they’re on the other side of the world, they generally have phone connections and they have access to a power grid. They sometimes have Internet and we can Skype, or we can call on a cellphone … they can send us things like video, we’ve helped them build their website … these kinds of things are possible, and it’s been nice to see their view of technology expand.
Some of our other communities are really difficult. For example we worked with a small school for the blind in Mongu, Zambia and that community just didn’t have any infrastructure. Even to call the school, you had to call the private cellphone of the principal or the headmaster, who might change over time. In places like that you have to find a third party … a person on the ground who can contact the school. You won’t get as frequent updates.
The biggest challenge is finding communication channels that work reliably and over time. Sometimes you won’t hear back until you get a chance to visit again, and that’s okay with us too. When we’re on the ground we try to do the best we can to not only create a technology of impact, but to also create and ecosystem of volunteers from the community who have some sort of technical background. We’ll train them in how to maintain whatever technology we’ve created … we try to build partnerships among different groups on the ground that might be beneficial to sustaining whatever technology solution we have developed, and then we hand it over to the community. And that’s okay.
With some communities we have more interaction, and with others, less. Language barriers can be a big problem unless you have a translator, but that’s okay too.
Audrow Nash: You’ve mentioned devices for the blind. What other devices and communities do you develop for?
Bernadine Dias: We’ve primarily focused on education. The first two years after we started our group were spent talking to communities around the world. We talked to anyone who would talk to us, whether it was remote or in person. We asked them about what they really wanted – not from technology, but in life – and their two biggest requests were for health and education. They wanted to be empowered to be healthy, and know enough to solve their own problems, to do their own thing, to earn their own living. After those two years, we were definitely more qualified to do education-related work, so that’s where we started and that’s been our primary focus since.
That said, we’ve done a few other projects, depending on what comes up.
Another popular project of ours looks at English literacy. A lot of communities want to improve their English literacy for job reasons.
You have teachers who want to improve their skills, but who sometimes haven’t been trained very well, who don’t have access to a lot of technology, who often work multiple shifts, and are so busy that they really don’t have time to do a lot of extras.
Then we have students … we’ve worked with a wide range all the way from primary school students and children with special needs, to adult learners, immigrant workers in Doha for example, and refugees in Pittsburg … who all have very different motivations for learning English and the kind of English they want to learn. Think of it as functional literacy versus educational literacy.
We were looking at ways to motivate students and teachers, and give them tools to work together. One exciting way to do this is through games, of course. There is a triangle of people … the game developers who want to build games that are cool, the teachers who have their own subjects and ways of teaching, and the students have to do homework but would rather be playing the games they are in to … so we put all these three together and created a framework where you can do your homework while playing a game.
The teacher can upload his or her exercises, and the game developers only have to follow a few different rules for embedding this educational content in their games. They can decide, for example, that you will do quizzes or exercises to get more lives or get extra tools to get through your adventure game. It’s up to them, but we can have certain gradations, so that if it’s a quiz then it has to happen so often, for example. Then the students get to decide whether they will play the Dragon Game or the Adventure Game or this role-playing game, etc … they get to pick their exercise and their game, and they get their homework done.
It’s been really successful because everybody gets to do what they’re best at: the teachers like it because they don’t have to do the extra work of changing the format of their educational materials; the students like it because they get to the play games they want; and the gamers like it because there is a social component to what they are doing.
We have more requests for this project than we can field for it, and we’re at the stage where we have a pretty good prototype and we’re doing field tests around the world.
Audrow Nash: In working with communities to develop these projects, what kind of lessons have you learned?
Bernadine Dias: So many! I think the biggest lesson that I have learned is how incredibly resourceful people are. We travel the world and we work with communities where I just cannot imagine what I would have done in that situation, and they come up with very brilliant ways to do things. We work the deaf-blind community for example … and when I’m trying to teach my 4-year-old son to read and write, or my 1-year-old daughter to point to things and when I say “this is a bottle” or “this is the TV,” all I can think is: if my child was born both blind and deaf how do I even do that?! It’s just incredible to me how people find ways.
I think the biggest lesson for technologist is that you have to be open to learning; you’re not the expert in anything other than the technology. When you go into a situation with that mindset, then you end up building much more impactful solutions. You end up having a much richer experience because of it, instead of getting frustrated about not having a data connection or having a power failure or whatever it is that comes your way.
Another lesson is learning to roll with the punches. You cannot plan everything. You have to plan a lot, there is a lot of logistical planning that goes on especially in field tests, but you have to be prepared to adapt at any given time.
The final important lesson is diverse teams. Getting people of different backgrounds who can come up with very creative solutions, whatever conditions get thrown your way.
I could talk about lessons for years, but it’s incredibly rewarding work, and there is never a problem recruiting people to work on these projects. People find them incredibly motivating. We handle the logistics and creating the partnerships, which most people find very difficult to do. Given the opportunity, people are very excited to work on these projects, and everyone is touched by them. You have stories to tell your grandkids for a lifetime.
Audrow Nash: How would someone get involved in helping these communities?
Bernadine Dias: You asked earlier about lessons learned, but the flipside to that are the challenges that continue to arise. Finances are the biggest problem. People in the US cost a lot of money, people in Europe cost a lot of money, and people in Australia cost a lot of money as bcompared to someone who lives a developing community. Just paying for people’s time and travel costs money, and not a lot people, institutions or organizations fund this work.
At the end of the day it all comes down to funding. We’ve tried to provide internship opportunities where we can. We’ve had people come in and work with us from other universities, but these are much more limited opportunities. We’ve also tried to mentor others who are interested in starting similar efforts in their own universities, and we’ve certainly helped a lot in answering questions about how we set things like students groups up. We’ve also connected them to other organizations in the areas that we might know about, where they can get a similar experience.
We also help make connections with partners. Sometimes we’ll get requests from partners in the developing world, but we don’t have the right technical expertise for whatever they want, and then we’ll try and connect them to some else who we know has an interest. That’s the best we can do right now.
What’s nice is that in the ten years we’ve been doing this, more and more groups have been starting to do this work – not only universities, but now companies as well … startups that have grown and are now taking interns and hiring people. That’s been exciting, and that’s probably something we’ll compile through our website now that more people are asking for that information.
Audrow Nash: We all know about the publish or perish paradigm … how can researchers get involved in doing this kind of work if it means not publishing as frequently?
Bernadine Dias: It’s less about not publishing as frequently, and more about not being able to publish in the venues that are considered esteemed. You can publish – there are a lot of lessons learned and there are a lot of things that come out of it, but it’s about what people expect and what’s traditionally viewed as great or not.
There are two methodologies for this …
One is that, as an academic community, we need to create more venues for people who do this sort of work to publish, where there is a community that understands and values the work, and that can meaningfully review the work and contribute to tenure cases and such.
We’ve trying to do that through a community called Information Communication Technology for Development (ICTD). It brings together social scientists, technologists and humanitarian workers to take courses together and to publish in joint venues. We found that the social sciences were much more amenable to this, and their cases were going through a lot more easily than the technology cases. There was an ACM DEV conference started in 2010 that started as an offshoot of ICTD and now it’s become a conference on its own … it’s purely focused on the technology side but all the people who come to that conference are working in this space, so they can give more technical feedback. I’m hoping that will continue to grow.
Meanwhile, people still have to figure out their lives until this all gets accepted. So another method is to carve out a niche within your space of expertise, and when you write your regular grants and you publish your standard publications, you push part of that work into the developing world or whichever underserved community you’re most interested in working with. In that way you can cross subsidize the funding.
On the publication side, there are the major publications, but then you also have [development-focused] publications too. They can give you good PR, and generally universities like that.
As long as you don’t do only this kind of work, it ends up being a benefit …
… Unless you want to be a renegade like me, and go all-or-nothing, which is fine too, if that’s who you are!
Audrow Nash: Thank you!
Bernadine Dias: You’re welcome!
All audio interviews are transcribed and edited for clarity with great care, however, we cannot assume responsibility for their accuracy.