App to determine water potability

17th March 2023 By: Leah Shelene Asaram - Features Reporter

South African charitable trust the Claude Leon Foundation is funding two research chairs, one of which will conclude research on a potable water test-strips application (app) to promote realisation of the right to safe water in vulnerable communities as well as citizen science.  

This research programme, which officially started on January 1 this year, will run for five years under the Water Stewardship programme and is worth R15.7-million.

The inter- and transdisciplinary programme of research, postgraduate supervision and advocacy is tackling current and future water challenges to find solutions that will benefit communities in South Africa, and eventually Africa.

The idea for the potable water test-strips app is to dip a test strip into water, photograph an image of it and upload it onto the app, which will then generate an answer to the question of whether the water is potable, and if not, what must be done to make it potable.

University of the Witwatersrand Centre in Water Research and Development (CIWaRD) director and Professor Craig Sheridan says the aim is that the app will use machine learning to execute colorimetry – using a phone with an app rather than the eye to interpret colour against a database of contaminant concentrations.

The team has recently recruited its first Master of Sciences in electrical engineering student to start working on the app.

He adds that the app-based project has also received cofunding from financial services company NinetyOne for research student bursaries and from the Water Research Commission to further develop the concepts.

Target Market

Sheridan expresses that, although the app is designed for those who reside in vulnerable communities, the app will be applicable for all those who are struggling to detect if water is potable or not.

“An app using the English language would be entirely inappropriate for roll-out in rural Eastern Cape. Equally, different users would want different data from the app.”

A crowd-science participant may want to know concentrations of different contaminants, as s/he might have a degree in civil or chemical engineering, whereas a hiker might simply want to know if the water is safe to drink or a farmer if the water is safe for their livestock or crops.

This outcome could be easily tailored to different groups of users, adds Sheridan.

He says that factors such as language barriers and literacy levels have been considered during the designing of the app and will form part of the development of the app – as opposed to development of the test-strips.

Cross-subsidising economic models is being considered and every strip would be disposable and environment-friendly.

Moreover, Sheridan says the app will run on data, but can still work offline and exchange data during app updates.

It will also work on any cellphone with a camera, and not necessarily the most up-to-date cellphone, therefore accounting for constraints of vulnerable communities.

Sheridan says the machine learning will be able to understand how every cellphone will interpret the colour of the test strip to provide a concentration of a specific contaminant.

There is still considerable development to be completed for the app with regard to the chemical, microbiological, engineering and legal aspects, which must be continued to reach the roll-out stage, he notes.

With the completion of all process stages, the app should be available to the market by the end of 2024.