Introduction
Just after five on a July morning, Theresa Riedl stood at the edge of a maize field and looked at two images of the same land: in front of her, a dusty furrow; on the tablet, a map in red, yellow and green. The red patch lay exactly where she knew an old drainage line crossed the field.
Theresa runs a family farm in Styria and now uses AI for irrigation, fertilization and the question of when a gut feeling is reliable enough. She prefers talking about soil rather than algorithms. It suits her.
Story of the Path into AI
The impulse came after several dry summers. Her son brought drone images that showed plant stress, and Theresa wanted to know whether there was more in them than a pretty image for presentations. The first systems seemed made for large, flat farms. Her fields are smaller, sloped and in places crossed by the shade of old fruit trees.
She learned to read soil sensors, compare maps and treat model suggestions as hints, not commands. Once the analysis mistook the shadow of a walnut tree for water stress; another time the digital field boundary did not match the actual embankment. Theresa wrote those errors into a notebook, beside weather rules her father had taught her. From that mixture came her first usable workflow: drone image, soil sample, farm memory, then decision.
Current Work
Today Theresa plans irrigation and nutrient applications with a simple dashboard. In one wheat section the system suggested an additional treatment because plant colour was uneven. Theresa checked the spot, found compacted soil along an old track and decided against a purely software-driven fix. Loosen first, calculate later.
The farm now saves some trips and some unnecessary measures. Theresa does not exaggerate this; a wet autumn can throw every plan out. Her main gain is an earlier view of problem areas before they grow large in the field.
Personal Advice
“AI is good at showing spots. Whether the spot has a history, you have to know yourself,” Theresa says. She does not mean that as hostility to technology. A person who knows her fields can use data better and notice faster when a model is only describing the surface.
Key Facts
Age and place: 52, Styria.
Background: family farm, climate pressure, practical agricultural experience.
Entry into AI: drone images and soil data for detecting stress zones.
Focus today: precision farming for smaller farms.
Typical tools: drone imagery, soil sensors, simple forecasting models.
Werkstattnotiz
Theresa still marks places where the model looks clean and the soil disagrees. Old interventions remain especially tricky: drainage lines, filled ditches, compacted vehicle tracks. She is testing which of these memories can be documented without pressing the whole farm into a spreadsheet.