Precision agriculture applies data-driven strategies to manage spatial and temporal variability within fields, aiming to increase productivity while minimizing pressure on natural resources. As interest in smart tillage systems expands, this study explores a central question: Can tillage tools be used to measure soil compaction during regular field operations? To investigate this, vibration data measurements were collected from a cultivator shank in the northeast of Kansas using the AVDAQ system. The test field soils were Reading silt loam and Eudora–Bismarck Grove silt loams. The relationship between shank vibrations, soil moisture (measured by a Hydrosense II soil–water sensor), and soil compaction (measured by a cone penetrometer) was evaluated using machine learning models. Both XGBoost and Random Forest demonstrated strong predictive performance, with Random Forest achieving a slightly higher correlation of 93.8% compared to 93.7% for XGBoost. Statistical analysis confirmed no significant difference between predicted and measured values, validating the accuracy and reliability of both models. Overall, the results demonstrate that combining vibration data with soil moisture data as model inputs enables accurate estimation of soil compaction, providing a foundation for future in situ soil sensing, reduced tillage intensity, and more sustainable cultivation practices.
Precision agriculture applies data-driven strategies to manage spatial and temporal variability within fields, aiming to increase productivity while minimizing pressure on natural resources. As interest in smart tillage systems expands, this study explores a central question: Can tillage tools be used to measure soil compaction during regular field operations? To investigate this, vibration data were collected from a cultivator shank using the AVDAQ system. The relationship between shank vibrations and soil compaction, as measured by a cone penetrometer, was evaluated using machine learning models. Both XGBoost and Random Forest demonstrated strong predictive performance, with Random Forest achieving a slightly higher correlation of 93.8% compared to 93.7% for XGBoost. Statistical analysis confirmed no significant difference between predicted and measured values, validating the accuracy and reliability of both models. These findings demonstrate the feasibility of using vibrations generated during tillage to estimate soil compaction under real-time field conditions. With further validation, this approach could be integrated into tractors to enable in-situ soil sensing, reduce tillage intensity, and support more sustainable and energy-efficient cultivation practices.
Vibration measured at the cultivator shank is a rich but underutilized source of information about soil–tool interaction, which we propose can be read as a language of the implement’s engagement with the soil. As a first step of understanding this language, this study focuses on the vibration signature of the turning phase of the tillage operation. During a turn, the shank is lifted out of the soil, the turn is completed, and then reinserted into the soil. Using a convolutional neural network autoencoder, trained with a custom dual-domain loss function on primary tillage signals collected in 2022 on silt loam, a shared pattern was extracted with a mean Pearson correlation of 0.965. This pattern was then validated against independent 2023 primary tillage data, yielding strong agreement primarily in the frequency-energy distribution (FFT r = 0.93, PSD r = 0.99), while time-domain and time–frequency agreement were more moderate. When applied to a 2024 dataset from secondary tillage on Chase silty clay loam, the broad energy envelope was retained, but the primary-tillage time–frequency distribution was not reproduced. The proposed framework provides a practical foundation for future vibration-based interpretation of tillage operations without requiring traditional terramechanics assumptions.