Projects per year
Abstract
Monitoring crops and fields is an important aspect in the agricultural sector to prevent droughts, floods or the spreading of insects and diseases. We introduce an intelligent solution to monitor agricultural environments. The system learns a model of the underlying field which it then uses to plan an optimal monitoring schedule. By interactively querying the preferences of the decision maker, we define a weighting to optimise over multiple objectives in the schedule, such as visiting frequency, intervention frequency and distance. We implement this as an interactive demo, called CropBot, using LEGO Mindstorms.
Original language | English |
---|---|
Title of host publication | BNAIC/BeNeLearn 2022 |
Publication status | Published - 9 Nov 2022 |
Projects
- 1 Active
-
FWOTM1082: Reinforcement Learning in Multi-Objective Multi-Agent Systems
1/11/21 → 31/10/23
Project: Fundamental