November 16, 2021
Development of an Analytics Platform for Real Time Monitoring of Water Pump Usage
Region/Country : Africa
contributor: Emmanuel Kinyanjui, Harsh Vyas
KickStart International commissioned the data logger project to enable the team to monitor, observe, and get insights on pump functionality and utilization. The logs retrieved from the pumps provide valuable insight on farmer behavior as well as a window for the technical team to remotely observe and diagnose pump issues.
The data loggers leveraged SMS (Short Message Service) messaging to send data to KickStart International’s Nairobi Office every five minutes. Using a message synchronization software, the team would then transfer messages from the phone to an excel sheet where the business intelligence team could graph out the data for insights on usage and pump diagnostics. A considerable challenge that emerged in this sequence was that the process was reliant upon bulk processing (at the end of the day or week) from phone to excel sheet, preventing the team from gathering real-time analytics to respond to issues quickly.
Over the course of the fellowship, an IoT Data Logger was developed to send data, in real-time, to a custom analytics service for immediate processing and visualization. The pumping devices were first configured to send data to ThingSpeak, an IoT analytics platform service that allows the user to aggregate, visualize, and analyze live data streams in the cloud. After this proof of concept was deemed successful, the team moved on to build out a custom analytics platform suited to KickStart’s specific requirements that would log the data, process it, and provide application programming interfaces (APIs) for a front-end web application to consume and visualize it for the KickStart team.
With this new approach the KickStart team now has access to real-time data on whether the pumps are functional, how much water has been pumped, and the pump’s real-time location. Furthermore, this has opened up the opportunity to tie the data to farm productivity which, alongside other parameters such as soil profile, can be used to boost the farm’s productivity to the benefit of the farmer.
Figure: The Dashboard
E4C 2021 PROGRAM MANAGEMENT TEAM: Mariela Machado, Senior Program Manager; Grace Burleson, Research Manager; Marilynn Holguín Clover, Program Coordinator; Jonathan Kemp, Program Associate
ADVISORS AND COLLABORATORS: James Mwaniki, Consultant Firmware Engineer, Kenya; Alan Sypbey, Director of Product Intelligence and Development, Kenya
This research was completed as part of the 2021 E4C Fellowship program. Learn more about the Fellows who worked on this research collaboration by connecting with them on LinkedIn: Emmanuel Kinyanjui and Harsh Vyas.