TERRITORY
Demeter IA
Software that helps to optimize irrigation management in agriculture, estimation of soil moisture, and makes accurate predictions about soil condition for the automation of irrigation processes.
SPAIN
AREA TO BE DIGITISED
Business Processes
SUBSECTOR #1
Agriculture
SUBSECTOR #2
Emerging crops
CROP PRODUCTION SYSTEM 1, 2, 3
Rainfed
Conventional
Open-field
TECHNOLOGY
Big Data and Data Analytics, Machine Learning
DIGITAL SOLUTION CATEGORY
Precision farming
STAKEHOLDERS
Agro-industries
IMPACT
Productivity
Demeter IA is a software that helps to optimize irrigation management in agriculture, estimation of soil moisture, and makes accurate predictions about soil condition for the automation of irrigation processes.
Farm challenges
Soil moisture measurement probes often lack advanced data analysis to optimize water consumption and fail to provide the uniform, daily irrigation that has been historically practiced. Satellite data moisture estimates have not enough depth and temporal resolution.
Innovative features of the initiative / solution
DEMETER AI is a software that can integrate sensoric data from the ground, satellite data and meteo with advanced data processing and machine learning techniques.
Results obtanied
1. Ecological: It would allow for predicting irrigation needs before they occur, delaying the event until it is truly necessary, with minimal stress for the plant.
2. Commercial: There is a growing demand for AI solutions in agriculture to optimize resource management, which Deduce Data Solutions can meet.
3. Economic: It would facilitate water savings, representing a positive impact on the local and regional economy.
Lessons learned
DEMETER demonstrates the potential of deep learning techniques in optimizing irrigation management in agriculture. The developed tool is capable of predicting irrigation needs for the next three days, representing a significant step towards more efficient and sustainable agriculture. However, to improve the tool in the future and explore possible expansions of its functionalities using satellite information, the following needs have been identified:
- The tool has been designed and developed for three types of crops (mandarin, corn, and pear), so it would be necessary to train it for other types of crops and with more historical data, including a greater number of annual cycles per location.
- It would be very useful to know the exact irrigation dates and the amount of water used in each irrigation to make recommendations, coinciding with the predicted time, as is common practice in the cited literature with similar solutions in other application cases.
- It would be desirable to have the coordinates of each individual probe for all locations for which data is provided. This would have potential benefits such as enabling initial attempts to increase the spatial resolution of a single probe.
Contact
- Website: https://deducedata.solutions/
