June 3, 2019
Fast, affordable water testing for the developing world
Engineering researchers combine biochemistry听and smartphone AI software听in new听system听
厂耻蝉丑补苍迟补听惭颈迟谤补
厂耻蝉丑补苍迟补听惭颈迟谤补 was doing field trials in remote areas of India when he was struck by how many people with听no听reliable access to听clean water were connected to the internet via smartphones.听
The contrast amazed him. It also gave him a good idea.听
Six years later, the University of 蓝莓视频 engineering professor and his colleagues are refining new technology that puts the power to monitor water quality in the palms of people in the developing world.
The system combines artificial intelligence (AI) software on smartphones with low-cost tools the researchers previously developed to test water for potentially deadly E. coli bacteria.听
听
Their ultimate aim is an affordable, comprehensive system to routinely monitor water for contamination without the delays and high costs听of听off-site laboratory tests.听
听
Woman collecting a听water听sample from a well.
鈥淚n remote locations, you鈥檙e talking about two days to take a sample, culture the bacteria and get results,鈥� says Mitra, executive director of the 蓝莓视频 Institute for Nanotechnology. 鈥淭hat is a major problem because people will keep drinking the water and fall sick or听even听die.鈥澨�
The new addition to听their听existing testing tools听鈥斕齱hich use paper strips, gel and plastic filters that change听colour听when E.听coli bacteria are present in water听鈥斕齣s AI trained to read听and听interpret the results.听
Users simply take a听smartphone听photograph of the strip, gel or filter and the AI application determines, based on听colour听change, whether听there is听contamination or not.听It has proven 99.99 per cent accurate, all but eliminating errors made by human interpreters.听
Work is ongoing to refine the system through training with more听data to also predict听E. coli听concentration, and therefore the level of contamination, by interpreting the intensity of the听colour听change.听
Woman collecting a听water听sample from a pump during field tests.
鈥淲e brought two disciplines together,鈥� says Mitra, a professor of mechanical and mechatronics engineering. 鈥淥ne is biochemistry and the other is a mathematical tool, a convolutional neural network.鈥澨�
Tests using mobile water kits听that were听first developed by the researchers cost about $15听each. A听subsequent,听simpler system using paper strips costs about $1 a test.听The AI app will be publicly available for free downloading.
Lab tests听to determine the degree of contamination, by contrast, cost about $60听apiece, an enormous expense in developing countries that lack听modern听water infrastructure.听
鈥淭his is a very empowering tool,鈥� says Mitra, whose primary collaborator is post-doctoral fellow Naga Siva Kumar听Gunda. 鈥淓ven people who don鈥檛 have access to clean water have access to cellphone networks.鈥澨�
In addition to countries such as India, the low-cost testing technology is attractive for use in remote or agricultural听areas听of the developed world,听and听for more frequent testing in modern municipal听water treatment听systems.听
Naga Siva Kumar Gunda (far left) and Sushanta Mitra (fourth from right) with officials during their field trials in India.
The new AI app could also be adapted for other kinds of tests听involving听colour听change. Possibilities include testing for heavy metals in water or diseases such as dengue fever and malaria.听
鈥淲ith this technology, we eliminate the need for independent lab tests,鈥� says Mitra. 鈥淚t saves time and costs as well. Our goal is to provide fast, accurate testing on a mobile platform听as technology for the masses.鈥澨�
听
