Our new generation Virtual Agronomist service uses artificial intelligence to communicate directly with farmers through WhatsApp and provide tailored nutrient plans and other essential agronomic advice for their plots.
So far, we have launched our AI chatbot service in Kenya and Uganda for maize and rice. Additionally, preparations are underway to launch Virtual Agronomist in Zambia for maize crops within the next few months. Virtual Agronomist supports the farmer with tailored advice throughout the season and can prompt and remind them of vital agronomic operations, for example, field mapping, soil data for their field, tailored nutrient plans, checking crop emergence, scouting for weeds, pests, and diseases, and harvesting. Through the features of our Virtual Agronomist service, farmers have the tools and advice they need to improve their crop health, yields and profits through better agronomic management.
But how do these features work?
Field Mapping
An example of the Virtual Agronomist field mapping feature
As part of the registration process for our Virtual Agronomist service, farmers are asked to walk the borders of their field and send their location at each of the corners. These location details allow us to provide a satellite image of their field along with an estimation of its size, knowledge that a lot of farmers have never had before. We then use this information to provide farmers with their tailored recommendations.
Soil Properties from iSDAsoil
Our iSDAsoil field-level soil map which offers open access soil data to everyone
ISDAsoil is the first field-level soil map for Africa, estimating over twenty soil properties across the entire continent. These properties range from chemical factors, like aluminium and pH levels, to physical factors, such as clay content and slope angles, and soil nutrients like potassium and zinc. Combining iSDAsoil data and field mapping allows us to get soil data estimations for the farmer’s whole field, giving them an insight into their soil health and fertility.
Tailored Nutrient Plans
An example of a maize nutrient plan
After registering their field, farmers can generate a nutrient plan. This plan shows them how much fertiliser, lime and manure to apply, and when to apply them throughout the season (basal fertiliser then topdressing) based on a target yield for that season. It also provides an estimate of the potential profit and costs farmers can hope to achieve for their yield.
The farmer answers questions from our AI chatbot about their field, for example, how much it has yielded in previous seasons, what yield the best farmers in the area are obtaining, and if the farmer is intercropping (cultivating two or more crops at the same time in the same field). A scientific model combines these answers with generalised local area data to suggest a target yield for the farmer’s field.
Knowing the farmer’s target yield and soil nutrient requirements from iSDAsoil, helps us determine the necessary fertilizers to enhance soil health and fertility. For this, a local fertiliser list is used, and the farmer can select which fertilisers are available to them.
The farmer can tweak the nutrient plan at any time to fit their needs. For example, a farmer may find that using three types of basal fertiliser achieves a better yield for less cost than if they just used one or two types of fertilisers. They can also adjust the target yield to be higher or lower depending on how much risk they are willing to accept during the season.
Pests and Diseases
Virtual Agronomist showing an example diagnosis from the pest and disease feature
The Virtual Agronomist pest and disease feature is a diagnostic tool which asks the farmer questions about how their plants are growing and the issues or symptoms that concern them, for example, insect larvae, lesions, leaf discolourations, or head smut. These descriptions are then passed onto our scientific model which provides the most likely diagnosis and suggests treatment actions that the farmer can take to combat the issues. The tool can also estimate the impact that the pest or disease could have on the farmer’s yields going forward if not managed correctly.
Seasonal Prompts
Virtual Agronomist showing some of the questions from the weed scout feature
Throughout the season, alerts are sent via our WhatsApp chatbot to farmers, reminding them to check crop emergence, to scout for weeds and to provide data after harvesting. To check crop emergence, the tool guides the farmer through the field, identifying any problems and suggesting solutions. Similarly, Virtual Agronomist helps the farmer assess weed growth and whether to weed or not to avoid yield loss.
At harvesting, the farmer is asked questions about their harvested yield and how they managed their farm throughout the season, for example, severity of pests and diseases and if they followed the advised nutrient plan. This allows Virtual Agronomist to suggest what might have impacted the farmer’s yield this season, and steps to improve their yields further for future seasons.
At iSDA, we are always looking to develop and enhance our Virtual Agronomist features in response to ongoing feedback from our users and will continue to refine our artificial intelligence technology. We believe that Virtual Agronomist will make it easier for farmers to access specific advice for their farms to help with all decisions they make throughout the season and improve the depth of agronomic guidance.
We have some exciting updates on the horizon for our Virtual Agronomist service, including developments for our pest and disease diagnostic feature using images and selection of crop varieties that have resistance to specific diseases. Keep an eye on our LinkedIn page for all our recent news and we look forward to giving more of an insight into the work we have been doing behind the scenes and on the ground to improve the livelihoods of farmers in Africa.
To learn more about Virtual Agronomist, click here.
