A Digital Advisor Twin for Crop Nitrogen Management

Abstract

Farmers and consultants face an unmanageable amount of diverse knowledge and information for crop management decisions. To determine optimal actions, decision makers require knowledge-based support. In this way, decisions can be improved and heuristics can be replaced over time. The study presents a digital knowledge base with an integrated decision support system (DSS), using the example of nutrient supply, specifically nitrogen (N), fertilization. Therefore, the requirements of farmers and crop consultants for DSS to inform fertilization decisions for winter wheat (Triticum aestivum L.) were elaborated using surveys, expert interviews, and a prototype test. Semantic knowledge was enriched by expert knowledge and combined in a web application, the Crop Portal. To map regional and personal decision making patterns and experiences, the tacit knowledge on the complex advisory problem of N fertilization is made digitally usable. For this purpose, 16 fuzzy variables were specified and formalized. Individual decision trees and their interactions with an integrative knowledge base were used to multiply the consulting reach of experts. Using three consultants and nine model farms from different soil–climate areas in Germany, the Crop Portal was tested under practical conditions and the perceived pragmatic and hedonic quality of the system was evaluated using a standardized questionnaire. The field test showed that the variation in fertilizer recommendations from the ‘digital advisor twin’ ranged from 5 kg N ha−1 to 16 kg N ha−1 when compared with the decisions of the experts in the field. The study presents the participatory development and evaluation of a rule-based DSS prototype in agricultural practice.

Potensi Penerapan Teknologi Digital Twin pada Industri Pertanian dan Pangan di Indonesia: Sebuah Tinjauan Literatur

“…Selain irigasi dan pengairan, juga terdapat pemupukan yang juga menjadi komponen penting selama masa pertumbuhan tanaman. Penelitian lain juga berhasil mengembangan model digital twin yang dapat memprediksi kebutuhan kuantitas pupuk (N) suatu lahan dengan memanfaatkan beberapa informasi data seperti cuaca, informasi jenis tanah, kelembapan tanah, dan suhu tanah (Weckesser et al, 2022;Kessler et al, 2021). Penerapan digital twin pada pemupukan juga dilakukan oleh Tsolakis et al (2019) dengan menciptakan robot simulasi otomatis yang dapat memberikan pupuk dan melakukan penyemprotan dengan terjadwal sesuai dengan pengaturan.…”

Digital Twins in the Automotive Industry: The Road toward Physical-Digital Convergence

“…The monitoring of land assets by installing various tools and sensors, i.e., in water, temperature, and humidity, for crop protection, maximization of profits, and reduction of failures [36][37][38][39][40][41].…”