20 Apr Akilimo: Taking cassava agronomy advice to scale
David Ngome and Pieter Pypers
The ACAI agronomy advice tool, Akilimo, was built to provide optimized and economically beneficial recommendations tailored to the biophysical and socioeconomic situation of cassava growers. The tool considers planting and harvest dates, local soil data, weather conditions, prices of available fertilizers, cost of land preparation operations, prices of cassava root produce, cropping objectives, risk attitude, and the investment capacity of the farmer.
ACAI has been conducting trials in Nigeria and Tanzania in collaboration with national research and development partners to find out how agronomic technologies affect the performance of cassava under different growing conditions. Results showed large variations in response to the variables, indicating the need for tailored recommendations for fertilizer regime, tillage operations, weed management, and advice on overall investments in a cassava cropping system.
Process based crop models and machine learning
To provide tailored recommendations, ACAI has developed an integrated system using machine learning techniques coupled with process-based crop models. To determine fertilizer recommendations, the ACAI team has combined the Light Interception and Utilization model (LINTUL), the Decision Support System for Agrotechnology Transfer (DSSAT), the Quantitative Evaluation of the Fertility of Tropical Soils model (QUEFTS), and economic optimizer algorithms to calibrate the recommendations. The mechanisms put in place determine the soil nutrient supply capacity, yield potential, nutrient-limited yield, and fertilizer rates required to acquire a target yield maximizing net revenue by combining observations from field trials, available GIS data, weather data, and the farmers’ ability to invest in fertilizer.
Using the QUEFTS model, the soil NPK supply was accurately predicted using the observed yield response in the Nutrient Omissions Trials. At these locations, the relationship between apparent soil nutrient supply and soil properties obtained from GIS layers from the International Soil Reference and Information Centre (ISRIC) was modeled using machine learning techniques. These models in turn were used to predict the soil NPK supply for the entire target intervention area. These soil properties can sufficiently explain the regional level soil variation. To explain soil variation at short range, however, the GIS layers need to be complemented with a local scale soil fertility indicator.
The use of common local soil fertility indicators, such as local soil name, soil depth/color, cropping history, perception of soil fertility, cropping history, manure/fertilizer use, etc., are not sufficiently generic as their predictive ability depends on the local context. Such indicators are therefore challenging to use in a standardized way. Within ACAI, current yield was found to be the best generic fertility indicator to adjust the soil nutrient supply at a regional scale to local soil conditions. This process forms a significant part of the research that has gone into developing the prediction engine.
Akilimo: Tailored cassava agronomy advice at scale
Akilimo is an all-in-one agronomic advisory tool that can be used to help intensify cassava farming and increase root and starch yields for cassava growers. The tool is currently providing tailored recommendations packaged in Interactive Voice Response, a smart mobile phone application and printable maps and guides. ACAI is in the process of developing infrastructure for short code formats as well as partnering with companies that offer integrated dissemination platforms like eSOKO and Viamo.
Akilimo combines initial use cases into a set of recommendation suited to the needs of the end user based on the input data and information requested. An extension agent will enter essential data inputs responding to a set of user-friendly questions into the Akilimo front-end application. The prediction engine running on a central server than receives a request and calculates the recommendation based on the received input data and sends out the advice by SMS, email, or directly within the application.
Akilimo is being integrated in the following formats: Smartphone app; Paper-based tools: flyers, manuals, lookup tables and maps; Interactive Voice Response (IVR); and Unstructured Supplementary Service Data (USSD).
The Smartphone App can be downloaded from Google app store. The app collects user-defined variables like location, land area, etc., asking questions in a stepwise manner. It interacts with the prediction engine on the server to provide recommendations directly on the smartphone or via SMS or email.
The Paper-Based Tools summarize the recommendations made using most logical values for input variables and present them using tables and maps. The paper-based tools are highly simplified versions providing simple best recommendations based on a limited set of input variables. For fertilizer recommendations, for example, best recommendations on a hectare basis are provided for each planting month per state or local government area (LGA).
The USSD approach requires the user to dial in a code which will let users answer several questions to define their location, resources, and other conditions and as in the smartphone app it interacts with the prediction engine to provide tailored recommendation via SMS. It is more versatile and site-specific than paper-based tools, but less so than the smartphone app. The USSD approach is highly favored by farmers who have simple feature phones and who wish to access the recommendations directly, without additional support from an extension worker.
The IVR method presents the same questions as in the smartphone app and USSD but in IVR, users can listen to the questions and select the answer from the provided choices.
One of the major challenges to improve the accuracy of the recommendations is the quality of the price data both for the fertilizers and the cassava roots. ACAI is exploring partnerships with various organizations providing digital market information as well as price mapping to provide meaningful default values.
Future steps include validating Akilimo both functionally—verifying whether the recommendations outperform current practices in the field, and architecturally—evaluating the user friendliness and how the tool can best fit within the dissemination strategy of development partners. ACAI is actively seeking interest from secondary partners to further test and scale the use of the Akilimo innovation.