It’s one thing for technology to work in the office or lab. It’s another whether it works properly in the field. Users have an annoying habit of getting in the way of the best laid plans.
Take water filters. There are now dozens of smart ideas aimed at the developing world, where millions still get sick from water-related diseases. But sometimes the theory and promise of these devices rides ahead of their impact. Sometimes, NGOs and others are better at distributing fixes than following up to see if they’re delivering.
In the name of accountability, a team at MIT came up with the CleanData-CleanWater sensor, a smart tap that fits to many filters, allowing NGOs to track whether and how they’re being used.
“We’re drowning in technologies to clean water. It seems like a new way comes out every day,” says David Taylor, a PhD candidate in mechanical engineering. “The problem is practitioners have no idea which one of these ideas work.”
The project grew out of a conversation with Susan Murcott, a lecturer in MIT’s Department of Civil and Environmental Engineering. Murcott runs an NGO called Pure Water Home, which has distributed clay filters to 100,000 people in Northern Ghana. Murcott wanted to understand more about usage and usability, and asked the students to come up with something.
Taylor, who worked in the region himself, explains that NGOs often get false data for their programs. A representative will go to a home and say “have you used that shiny new device we gave you?” To which, a user will invariably say “yes”–even if it’s not true. “People feel obligated to say ‘yes.’ It’s like when the dentist asks you if you’ve flossed recently. Everyone says yes,” Taylor says. He reckons, typically, the data might be off by as much 25%.
The smart tap, which is designed for water cooler-type faucets, is made up of a microcontroller, sensors, and battery, which lasts for up to a year. It records when the filter is opened and for how long it’s used for, down to the second. At the moment, it costs about $10 a unit.
“It allows the NGO to know whether they should invest in more water filters or whether they should invest in training the people to use water filters,” Taylor says. If less water is passing through than might be expected for a whole family, it may indicate the device isn’t easy to use, or that certain family members aren’t happy using it.
Back in May, CleanData-CleanWater won the $10,000 first prize in MIT’s IDEAS Global Challenge. It’s now using the money to fund production of the first 1,000 sensors, which will go to Ghana before the end of the year. Eventually, Taylor would like to set up a business selling hardware and services to NGOs, generally encouraging a more data-driven approach to water filtration programs.
“By getting the cost [of the sensor technology] down, this can become a metric for whether projects get funded,” he says.