Since 2019, Kinara for Youth Evolution has been successfully using mWater in our “Water for Life” project to enable our field volunteers, or “Community Change Agents” (CCAs), to identify and help fix water pipe leaks in Morogoro, Tanzania in partnership with the Morogoro Urban Water Supply and Sanitation Authority (MORUWASA) with the aims to reduce non-revenue water and increase household water availability. Dedicated to continuously improving our operations and advocacy efforts, we take a look at mWater’s new Population Data feature as an added dimension of live, localized data.
Kinara for Youth Evolution’s “Water for Life” project completed its second project phase, supported by Innovation for Change (I4C) Africa Hub, in May 2021. In the past six months, we scaled-up our work from 10 to 15 wards within Morogoro Municipality, fixed 430 leaks, resolved 101 complaints, and tracked household water availability through monthly surveys in 10 households per ward. All of this data is collected by CCAs, communicated to MORUWASA, and managed by Kinara through the mWater platform. While aggregate numbers are easily computed with mWater’s Portal tools, it can be a challenge to prioritize our work and advocacy with upwards of 100 open leaks at any one time, and 150 new household data points per month. It is common knowledge that certain wards have a high population density and thus a pipe leak, especially of a main connection, would affect many more people than in less populated areas. mWater’s new Population Data feature, which combines regularly updated satellite imagery, census data, and advanced machine learning techniques, is a promising enhancement to our ability to make such decisions because it provides data at a localized, programmable granularity.
The map and table in Figure 1 shows an example snapshot of open leaks together with population density data. In this example, Mazimbu Ward would seem to be a priority area based on both the number of open leaks and the large number of people who may be affected. Other wards are more mixed so a nuanced discussion with MORUWASA is needed to prioritize between high-leak wards with lower population densities (e.g., Mji Mpya) and wards where even a few leaks may affect many more people (e.g., Mafiga). Other important variables include the number of days that each leak has remained unresolved, the amount of water actually flowing, and whether this is a main connection affecting an entire neighbourhood or an individual household connection. Looking at historical data of all leaks discovered can help us plan how volunteers and technician plumbers can be better allocated for future work.
Figure 1: Open leaks example map and pivot table with population data
An example of surveyed households with population data is similarly shown in Figure 2. MORUWASA does well to balance available water days for the higher density wards (average "within 200 metres" population greater than 1,000) which receive above or just below the overall average. The middle and lower density wards show more mixed results. Many factors could contribute here including new infrastructure for growing populations, distribution policies itself based on population, and seasonal variations. As with leaks, the data can be used to drive further collaborative discussion with MORUWASA and help suggest improvements such as reviewing water distribution strategies.
Figure 2: Surveyed households example map and pivot table with population data
Both maps above visualize the established high density wards northwest of town centre, such as Chamwino Ward where Kinara is based, and the less populated but growing areas particularly to the north on the highway (marked B129) towards the capital city, Dodoma. Looking to scale-up our solution further, population data can help us identify growing areas of Morogoro Municipality and beyond to other towns and cities in Tanzania, which cross-referenced with information from local authorities and volunteers, gives us a fuller picture of how we can best expand our project’s impact.