The project explores the development of remote sensors and mobile technology to monitor use and functionality of pumps to predict failure before it happens.
80% of the 1.3 billion people using boreholes to access safe water live in rural areas. In these areas, hand pumps are the preferred technology, as they can lift widely available and good quality groundwater at low cost. However, up to one third of hand pumps do not function at any one time across Africa, with many abandoned by their communities within 1-2 years of a projected 10 year plus lifespan. A survey of users in Kyuso, Kenya finds the cost of repairs (19%), time taken to carry out repairs (17%) and frequency of breakdown (17%) to be leading factors for dissatisfaction with hand pumps. By contrast, only 1% of users complained about the fee for water usage (Oxford/RFL, 2014).
There is currently product development and research around the use of remote sensors and mobile technology to monitor use and functionality of the pumps and alert governments or organizations to potential problems in real-time. UNICEF WASH and many other partners are already piloting this technology in pilots in the field and learning about ways to leverage the data and information gathered.
While this real-time data offers the potential to improve response time for pump repair, there will still some down-time for the pumps. Most service agreements require the private sector to repair within 48-72 hours from when a sensor indicates failure. Thus, it’s possible that communities will not have access to clean water for at least 2-3 days.
In an attempt to address this issues, UNICEF is interested in exploring additional technology that can be integrated into the new monitoring devices to predict pump failure, in order to give local governments and organizations more time to plan and respond.
Predictive technology is fairly common in today’s more hi-tech fields such as airplane engines, medical devices, and computer hard drives. The technology is based on statistical analysis of historical data (mostly mechanical) in order to develop a model or algorithm that can predict malfunctions based on patterns and harmonics.
The project has two main goals to:
1. Develop predicative failure algorithms specific for hand pumps (laboratory proof of concept).
2. Field test the technology as part of a larger sensor/mobile real-time monitoring project (field proof of concept).
This technology has the potential to improve sustainable access to clean water. Predictive failure has not previously been considered for hand pumps typically installed by UNICEF and its partners. However, there is already momentum in the sector for the use sensors and mobile technology to monitor the use of hand and foot pumps for boreholes in rural areas. This project aims to capitalize on that momentum and, by supporting the proof of concept ground-work, ensure that predictive technology is developed and validated so that it can be integrated into the development of these products.