Adapting smartphone app used in water testing, for soil nutrient analysis

Abstract Smartphone technology has now penetrated every aspect of modern life. At such high rates of access and utilization, there is today much potential for the development of smartphones as high-performing tools in a number of industries. Traditionally, smartphones have been used as e.g. point-of-care testing devices in developing countries; now a similar approach can be extended to agriculture. This paper assesses the viability of utilizing smartphones in soil analysis. An Android-based smartphone application, in conjunction with commercially available Quantofix® test strips, was employed to analyze 92 soil samples collected across Indonesia. The soils tested encompassed a wide range of different textures (with 13%, 60% and 25% of samples constituting sandy, loamy and clayey soils, respectively), soil organic matter contents (range: 0.8–19.7%) and nutrient concentrations (range for plant-available N: 0.1–137.4 mg kg−1 and P: 1.2 to 64.2 mg kg−1; on dry soil basis). The app utilizes the smartphone as a portable reflectometer, which relates the color of test strips to the concentration of particular nutrients present in the soil medium. Three mobile devices currently available on the market, representing low, mid- and high-end products, were used to test the application. The results obtained via the smartphone were compared against standard methods for determination of extractable nitrate-N and exchangeable phosphorus (Olsen-P) under laboratory conditions. The smartphone-mediated soil analysis was found to have a high degree of agreement with standard methods for nitrate-N determination (87% of samples with nitrate-N differed by less than 10 mg kg−1 from the standard method for the high-end smartphone) but not for phosphorus determination where chemical interferences to test strip colour development were noted. All three mobile devices were shown to be effective as portable reflectometers. However, color perception was found to differ amongst the devices, resulting in a consistent bias between the high-end phone and the remaining appliances. Whereas, it is essential to consider the inter-smartphone variability in readings and environmental factors such as temperature prior to the smartphone-mediated soil analysis, the smartphone-test strip combination might be employed as acceptable screening tool for soil nutrient concentration assessment to enhance crop outcomes, increasing yield, and preventing over-application of inputs, reducing consequent financial and environmental impact. Further enhancements can test the applicability of smartphone-mediated soil analysis in field conditions.