AgriRover was developed to address key operational constraints faced by smallholder vineyards in Peru, including sandy and saline soils, labor shortages, and limited access to advanced agricultural machinery. The platform features an articulated, all-wheel-drive chassis designed to ensure mobility and stability on loose terrain while minimizing soil compaction. This study presents the simulation-driven development of a digital pre-twin, conceived as a virtual prototype prepared for future sensor integration but currently operating without real-time data feedback. The pre-twin was implemented in MATLAB/Simulink (vers. 2024b) using a multibody dynamics model and evaluated through eight scenario-based simulations, varying field geometry, soil type, and slope conditions. The results show stable operation on slopes up to 10°, wheel sinkage values ranging between approximately 20 and 45 mm depending on terrain conditions, and a moderate battery state-of-charge reduction across most scenarios, with higher power demand observed on sandy soils. A scenario-based comparison indicates a potential reduction of approximately 50% in total monitoring time relative to manual field scouting, while advanced sensing, autonomous navigation, and AI-based analytics remain part of future developments. The current pre-twin provides a validated, low-cost foundation for context-specific phenological monitoring and early-stage precision agriculture applications in developing regions.

Tracking structural behavior is critically important to reduce maintenance and repair costs. Structural Health Monitoring (SHM) aims to evaluate the structural integrity, detect damage or abnormalities, and estimate overall safety. The integration of Machine Learning techniques has significantly advanced SHM by enabling the identification of deterioration patterns through sensor data analysis. This study focuses on classifying different vibration patterns recorded under various excitation scenarios (ambient, transient, and forced) using sensors installed directly on a 3-DoF structure. The proposed approach used a two-dimensional convolutional neural network (2D-CNN) trained on vibration image patterns generated from vibration signal scalogram images. To address dataset imbalance, stratified 5 × 3 Nested cross-validation and multiple performance metrics were computed to ensure robust evaluation. The proposed method was compared with single-sensor scalogram approaches and baseline models, including Support Vector Machines (SVM), Random Forest (RF), eXtreme Gradient Boosting (XGBoost), One-Dimensional Convolutional Neural Network (1D-CNN), and Long Short-Term Memory (LSTM) models, incorporating class-weighting strategies. Additionally, the contribution of the Total Energy Delivered by Sensor (TES) feature was evaluated for SVM, RF, and XGBoost models. The 2D-CNN model achieved superior performance in identifying excitation types associated with structural dynamic behavior, highlighting its effectiveness for structural vibration pattern recognition in SHM applications.