A smartphone-based early alert system for screening of coliform contamination in drinking water

attached to a smartphone. Current methods of coliform detection rely upon relatively expensive laboratory-based time consuming
techniques which require trained manpower and take at least 24-48 hours. This waiting time prevents quick action and the consequences
can be severe since the contaminated water may already have been consumed by then. Instead an unattended smartphone can
continuously monitor the sample and send an alert as soon as contamination is detected. Smartphones, especially older or unused ones,
fitted with a customized compact incubator and a sample holder, can be set to take photos of the sample (mixed with a selective growth
medium) at regular intervals. An image analysis algorithm would analyze the photos and predict contamination as soon as it notices any
increase in turbidity and/or change in color of the sample under observation due to bacterial growth. On detection of contamination,
alerts can be immediately sent out to the concerned parties and intervention can be made without any potentially harmful delay. To test
this concept we built a prototype for the detection of coliform/E. coli contamination in water samples. With the initial bacterial counts
varying from 1-10 to >108
colony forming units (CFU) per 100 ml of water samples, all the results were produced within a turnaround
time of 4 to 12 hours and found to be comparable with conventional microbiological methods which require 24-48 hours of incubation.