Data gathering usually requires an extensive infrastructure, but open mobile technologies could change that. An interview with Yaw Anokwa of the Open Data Kit project.
Why did you choose to focus on improving data collection in the developing world?
Our research group has historically worked on using embedded systems like phones to sense humans and their environment. Over the past few years, we’ve moved into using sensors to deliver better and more standardized healthcare.
The focus in developing regions stems from personal interests from a number of people in our group. Unlike other researchers on technology in developing regions, we primarily work on what computers can do well in developing regions, which is optimizing existing processes. Data collection on paper is one of the most common and inefficient processes, and with the growth of mobile technology in these areas, using phones for it is a natural fit.
The Open Data Kit puts together a number of technologies, including Google’s Android mobile phone platform and App Engine. What requirements guided your choice of technologies?
The biggest reason we use Android is because it’s open source. It’s not the first open mobile operating system, but it’s the first open and comprehensive system with the wide and viable device support that our users need.
Moreover, as researchers we need to be able to transform the mobile device into whatever we imagine, whether it’s a mobile ultrasound unit or a gesture recognizing computer. Android’s open source framework means we are free to innovate however we see fit, independently of what other organizations in the Android ecosystem do.
The choice of App Engine is driven by the use cases we’ve seen in the field. It’s difficult for most organizations in resource constrained regions to keep infrastructure running. Simply keeping the power running and the data servers up to date can be quite challenging.
For the many places that have connection to the web via the cell phone network, it is often more time and cost effective to use something like Google’s App Engine. Another way to think of it is by analogy with the email infrastructure. Running an email server is hard, but with services like Gmail anyone can have email. We want to build services that make data collection as easy as email.
What feedback have you received so far from potential users? How do you see the Open Data Kit being used in the future?
Users really like ODK. Even though it has only been out for a few months, we are seeing wide adoption. We have organizations using it to triage and diagnose patients in Tanzania, and one using it to document war crimes in Central African Republic. Those organizations find that it’s really easy for non-experts (some of whom have never seen a phone) to pick it up and use it. The fact that it supports multiple languages, input constraints, barcodes, and picture, audio, and video capture and playback also makes it very attractive.
Some of the more exciting research ideas we are working on include connecting external sensors to enable field medical devices like ultrasound and stethoscopes. Once captured, the data can be analyzed on the phones and some basic automated diagnoses can be performed.
Beyond the medical domain, there is also great citizen science that can be performed. For example, a city can use ODK to gather data from their citizens. The city could publish forms that gather information about pollution or potholes or political feedback. That data can include images and GPS locations and all that could be submitted by citizens. The city could get instant feedback about all these issues.
The real power of ODK is that we don’t control its future. We’ve already seen a number of organizations taking the code and building their own tools on top of it. We’ve also started teaching a class at the University of Washington where graduate and undergraduate students are extending the functionality. As the project grows, we hope more people add to and change ODK for their own needs but always with a bias towards supporting the larger ODK community.
IEET Blog