Physical services
Overview
In the apple orchards of Italy, precision and timing are critical to ensure quality, profitability, and sustainability. To support farmers in managing increasingly complex harvest operations, innovative start-up GeoInference developed an AI-driven solution that redefines how apples are counted and sized during harvest.
With the support of agrifoodTEF and the agrifoodTEF Italian Node, GeoInference tested and optimised its solution in real operational conditions. By combining AI-powered vision, real-time monitoring, and data analytics, the system enables more accurate harvest management while integrating smoothly into daily farming activities, even in environments with limited connectivity.
Impact
Thanks to testing and support provided through agrifoodTEF, GeoInference’s solution delivers tangible benefits for apple producers:
- Enables real-time counting and sizing of apples, increasing accuracy compared to manual processes.
- Improves harvest planning and production forecasting, supporting better decision-making.
- Reduces human error and operational inefficiencies, saving time and resources.
- Helps minimise waste and environmental impact through more precise harvest management.
- Demonstrates adaptability beyond apples, highlighting potential for broader application across crops.
The results confirm the solution’s reliability and relevance for modern, data-driven agriculture.
The Challenge
Accurate estimation of yields is essential for effective harvest planning, yet traditional methods such as manual counting and fruit calibration are time-consuming, error-prone, and increasingly inadequate for modern agricultural demands.
Apple producers also face growing uncertainty due to variable weather conditions and competitive global markets. To remain sustainable and profitable, farmers need solutions that enhance resource management, complement existing practices, and provide reliable data without disrupting workflows.
The Solution
GeoInference developed a patented AI-based system that automatically counts and sizes apples during harvest using advanced vision technologies. Each harvest batch is assigned unique data tags, allowing real-time tracking of fruit size and quantity.
With the support of the agrifoodTEF Italian Node, the solution was tested and fine-tuned to ensure robust performance in real-world conditions, including low-connectivity environments. Harvest data is uploaded to the cloud, enabling farmers to easily monitor yields and evaluate performance.
Designed to integrate seamlessly into daily operations, the solution enhances efficiency without requiring major changes to existing processes. By enabling real-world testing and validation, agrifoodTEF supports the deployment of AI technologies that help farmers transition towards more precise, sustainable, and competitive agricultural practices.
Ready to test AI-driven farming solutions in real agricultural environments? Discover how agrifoodTEF turns innovation into field-ready impact.
