SMARTOLIVAR360 Working Group: Predictive Performance Modeling of the Olive Grove Sector using Augmented Intelligence

A predictive model for olive grove performance has been developed and trained using real-world data (productivity, climate, soil, management) capable of predicting olive and oil production per plot with a low error rate in validation tests. This model will incorporate advanced algorithms (e.g., neural networks) and will be adjusted to different types of olive groves (dryland, irrigated, variety, planting density). This development aims to bridge the gaps that persist in current innovations applied to olive groves. Existing solutions have focused on prediction (what will happen), but they lack the ability to generate personalized agronomic recommendations on what to do in response.

A prescriptive agronomic recommendations module is integrated into the model, translating predictions into specific management advice (irrigation, fertilization, phytosanitary treatments, pruning) for each farm. The recommendations are based on decision rules and machine learning, suggesting concrete actions (for example, adjusting irrigation based on the weather forecast or bringing forward the harvest on certain farms) to improve the expected productivity or quality. This proposal represents a qualitative leap from passive data prediction to proactive action. Current tools can predict harvests or risks (pests, adverse weather), but they don't indicate "what to do" next. SmartOlivar360 closes this gap by automatically and scientifically generating personalized agronomic guidelines (related to irrigation, fertilization, pruning, pest management, etc.) to improve yield. No AI solution applied to olive groves today offers this additional step of prescription, which makes SmartOlivar360 a pioneering project in translating data into practical decisions for the farmer.

The "SmartOlivar360" mobile application has been developed with an intuitive, multilingual interface, allowing farmers to consult harvest predictions for their plots and receive real-time alerts and recommendations. The app will feature simple visual elements (traffic light indicators, suggested task calendars, and basic maps), ensuring its usability even for users with limited digital experience. The project emphasizes a simple and intuitive user experience. The application and its interfaces will be developed using rural user-centered design principles, validating prototypes with real farmers to ensure the tool is understandable and useful in the field from day one. Information and alerts will be presented in a clear and practical manner.

An agronomic messaging system (multichannel chatbot) has been implemented, operating via WhatsApp/SMS for users who do not use advanced smartphones. This chatbot will send personalized alerts (e.g., high risk of pests in the area with treatment recommendations) and answer frequently asked questions from farmers (such as, "When should I irrigate this week?" with answers based on soil moisture data and rainfall forecasts). This expands the accessibility of the innovation to any mobile phone, overcoming the connectivity gap. The system will be able to use traditional channels (such as SMS or WhatsApp messages) and will employ affordable plug-and-play devices. This approach allows any olive grower, even in remote rural areas or with limited resources, to benefit from the tool without technological or economic barriers.

A low-cost, integrated agro-climatic sensor kit that connects plug-and-play to the platform. It includes prototype wireless sensors (powered by solar energy or long-life batteries) to measure critical parameters in the field (soil moisture, ambient temperature, and tree nutrient status using simple foliar sensors). The sensors will automatically synchronize with the app (via Bluetooth or LPWAN) and their real-time data will feed the predictive model, increasing the local accuracy of predictions and recommendations. This will allow for modular scalability of the solution: from farmers using only satellite and average data to those incorporating sensors for personalized recommendations on their farms.

Pilot validation of the AI system in real olive groves. Results reports will be generated demonstrating the improvements achieved (for example, an X% increase in the accuracy of the harvest forecast compared to traditional methods, or Y% reductions in inputs thanks to the recommendations), verifying the model's effectiveness and refining its parameters with feedback from each season. From its inception, the project has been designed for scalability and broad impact on the olive oil sector. The solution will be based on resources available throughout Spain, so it can be easily replicated beyond the pilot region of Andalusia to other olive-growing communities.

A training program and pilot demonstrations were conducted with the participation of olive-growing cooperatives and local associations. At least 100 farmers with diverse backgrounds will test the platform in a pilot phase, receiving practical training to interpret predictions and apply recommendations. Their suggestions for improving usability will be gathered. This initiative aims to ensure the effective adoption of the technology by training end users and building trust in the tool through local success stories.

Programming of the dissemination action plan. This is the set of activities that comprise the operational group's overall dissemination strategy. This strategy includes the development of the corporate identity manual, the website, press releases, articles disseminating results, and all other communication activities, including dissemination events. This work will be carried out at a national level, as we believe that promoting the olive oil sector should encompass the entire country. All of this will be done while also recognizing the importance of holding events in the main areas where olive groves are located. Specifically, activities will be held in Madrid, Andalusia, Aragon, and Castilla-La Mancha.

The creation of audiovisual materials is an important part of disseminating the project and its results, and today it is one of the most popular methods with the general public. The audiovisual material generated in this project is in an ideal format for adding content to the project's corporate website (and partners can also include it on their respective websites), as well as on the social media of all GO beneficiaries. To maximize the reach of the content, we will use paid media, which will enhance our ability to reach a wider audience.

Evaluation of the reach of dissemination activities, which aim to provide concrete data on the true impact of communication initiatives. The scope is national, as these reports are intended for the administrations evaluating the implementation of project activities. Availability on the corporate website also means that the information is accessible throughout the country. The information will be hosted on the website so that it is available to anyone who wishes to consult it. The verification sources (KPIs) will be: KPI reports, digital publications (number of recipients), website (number of visits and users), seminars (number of attendees, photographs), and impact reports.