Intelligent Agricultural Tracked Mobile Robot for In-Situ Soil Macronutrient Profiling and Precision Decision Support
Year: 2026
Type: preprint
Abstract: Traditional agricultural soil diagnostics rely heavily on manual core extraction followed by centralized chemical laboratory spectrometry, imposing multi-day latency that impedes time-critical fertilization decisions. This paper presents the mechanical design, embedded sensor telemetry, and offline decision-support architecture of an autonomous-ready tracked Unmanned Ground Vehicle (UGV) developed for in-situ soil diagnostic profiling in uneven semi-arid terrains. The mobile platform incorporates an 8-arm independent spring-loaded suspension crawler chassis (408 mm x 289 mm x 132.5 mm) driven by high-torque DC gearmotors via dual BTS7960 H-bridges, ensuring high traction without track derailing across irregular furrows. An actuated vertical rack-and-pinion stage deploys an industrial 7-in-1 electrochemical probe directly into calibrated topsoil depths, acquiring Nitrogen (N), Phosphorus (P), Potassium (K), pH, electrical conductivity (EC), moisture, and temperature via RS-485 Modbus RTU with CRC verification. Spatial georeferencing is integrated via an onboard NEO-6M GPS receiver, while an ESP32 dual-core MCU coordinates local telemetry and visual feedback via an ESP32-CAM module. An edge Python/Streamlit agronomic expert system evaluates logged spatial fertility matrices against localized crop requirement thresholds, automatically synthesizing variable-rate fertilizer dosages. Field trials conducted on unpaved agricultural soils in Aleppo demonstrated robust obstacle negotiation (>30 deg incline) and high electrochemical repeatability (<4.1% relative deviation against standard laboratory bench tests), demonstrating a cost-effective paradigm for precision nutrient management.
Authors: Esleem, Abdulkader, Jannat, Yaman
Institution: University of Aleppo
Language: English
Topic: Smart Agriculture and AI
Subfield: Plant Science
Domain: Life Sciences
Open Access status: green
OpenAlex ID: W7224127573