This blog post is based on the 29th Practice Abstract from the second batch of Farmtopia Practice Abstracts.
For small fruit farms, decisions about irrigation and fertilisation are often based on experience, regular field checks and weather conditions. But when conditions change quickly, it can be difficult to know exactly when and where trees need water or nutrients.
Smart4Plums, developed through the Farmtopia Open Call, uses soil sensors, weather information and AI-powered decision support to give plum growers a clearer picture of what is happening in the root zone and help them make more precise irrigation and fertilisation decisions.
The challenge: knowing what is happening below the surface
A walk through the orchard can tell a farmer a lot, but it cannot always reveal what is happening in the soil around the roots. Periods of heat or dry weather can quickly change soil conditions, while water and fertiliser applied away from the root zone may provide little benefit. For small farms with limited water supplies and tight budgets, this unnecessary use can be particularly costly.
In plum production, timing is also important. Applying water or nutrients too early, too late or in the wrong quantity can affect crop quality and yield. Smart4Plums addresses this challenge by continuously monitoring conditions in the root zone and turning multiple measurements into practical information for farmers.
From soil measurements to practical advice
The system uses a small number of strategically placed soil sensors to measure parameters including:
- soil moisture;
- temperature;
- acidity (pH);
- salinity;
- and nutrient levels such as NPK.
The data is collected through a local system and displayed on a dashboard that can be accessed from a phone or computer. Weather information is also incorporated, helping farmers interpret soil measurements in relation to current and upcoming conditions. Instead of looking at each measurement separately, the D2Port AI framework analyses the parameters together. The system uses colour-coded charts to show the overall behaviour of the orchard over daily, weekly and monthly periods.
This makes it easier to identify when conditions are stable, when they are beginning to change and when intervention may be needed.
Identifying the cause, not just the problem
One of the key features of Smart4Plums is its ability to support root-cause attribution. For example, an alert does not simply indicate that something is outside the expected range. The system can identify a combination of factors contributing to the deviation, such as low soil moisture together with high temperature and declining nutrient levels.
This helps farmers decide whether irrigation or fertilisation is actually needed and where it should be targeted. The system can also distinguish between normal seasonal changes and potential anomalies, helping to reduce unnecessary alerts while still highlighting conditions that may require attention.
Tested on a real plum farm in Serbia
Smart4Plums was not developed only as a laboratory concept. The system was installed and tested on a working plum farm in Serbia. The participating farm was equipped with:
- a water well and pump;
- drip irrigation;
- a Venturi fertiliser injector;
- long-range Wi-Fi;
- and a local server.
The field testing demonstrated that continuous monitoring does not require sensors in every row. A limited number of well-positioned sensors in representative areas can provide useful information about soil conditions across the orchard. This also supports a more affordable approach for small farms: start with a basic monitoring setup and expand it as the benefits become clear.
What can farmers gain?The main advantage is a shift from irrigation and fertilisation based mainly on habit or fixed schedules towards decisions supported by current field data.
Smart4Plums can help farmers:
- monitor soil conditions continuously;
- identify changes and trends earlier;
- target water and nutrients towards the root zone;
- reduce unnecessary application between rows;
- respond more quickly to dry conditions or nutrient changes;
- combine local observations with sensor and weather data;
- and gradually expand their digital monitoring system.
The Practice Abstract gives an expected yield of approximately 40 t/ha, based on an expected production of 8 tonnes on 0.20 ha. This corresponds to an expected increase of approximately 60–100% compared with the baseline range.
Fruit quality is also expected to reach or exceed 16% Brix, compared with a previous farm-level reference of 14% Brix. These figures are presented as expected results, rather than measured impacts demonstrated across a full production cycle.
Practical recommendations for plum growers
Farmers considering a similar approach should:
- Start with a few representative locations rather than installing sensors across every row.
- Place sensors where they can provide a useful picture of root-zone conditions.
- Combine sensor information with weather data and field observations.
- Look at trends over time rather than reacting to a single measurement.
- Use the system’s analysis to identify the likely causes of changes in soil conditions.
- Target irrigation and fertilisation towards the root zone whenever possible.
- Start with a basic digital setup and expand it gradually as the value becomes clear.
The most important step is not simply collecting more data. It is using the data to recognise changes early and take action at the right time.
Bringing precision decision support within reach of small farms
Smart4Plums shows how AI and relatively simple sensor infrastructure can provide small fruit farms with a continuous picture of what is happening below the surface. By combining soil measurements, weather information, irrigation infrastructure and AI-based analysis, the system can help plum growers move from “Is the orchard dry?” to a more useful question: “Where is the problem, what is causing it, and what should I do next?”
For farms facing limited water availability, rising input costs and changing weather conditions, this kind of targeted decision support can provide a practical way to improve resource management while building digital capacity step by step
About this Practice Abstract
Official title: An AI-Powered Decision-Support System for Precision Irrigation and Fertilisation in Plum Orchards (Smart4Plums OC SIP)
Publisher: NISSATECH Innovation Centre, Farm DB
Authors: Nenad Stojanović; Darko Burdžić
Contact: nenad.stojanovic@nissatech.com; darkoburdzic@gmail.com