Mango firmness is one of the critical indicators for assessing internal quality and taste, as well as an indirect measure of maturity and freshness during ripening. Acoustic vibration technology has been widely applied for nondestructive detection of fruit firmness. However, existing detection systems face the risk of fruit damage, prediction performance limitations, and significant influence of fruit size. This study designed a nondestructive pneumatic-electromagnetic-driven impact device based on acoustic vibration technology for firmness detection of different sizes of mango with the same impact force control. Vibration signals of 156 mangoes were acquired using an embedded accelerometer, and effective vibration signals were selected by comparing the excitation vibration response signals and the free vibration response signals. The correlation between mango reference firmness and vibration signal features was then analyzed. Based on this analysis, a prediction model for mango firmness was developed using partial least squares regression based on competitive adaptive reweighted sampling (CARS-PLSR). The results showed that the energy-type and amplitude-type statistical features in the vibration signals had a good correlation with the reference firmness ( $\left|r\right|$≥0.45), and the mango firmness prediction model based on the vibration frequency-domain signals (CARS-PLSR) had the optimal performance. The model’s prediction determination coefficient ( $R_P^2$), root mean square error of prediction (RMSEP), and relative percent deviation ( $RP{D}_P$) were 0.95, 0.29 N/mm, and 4.20, respectively. Overall, it demonstrated that the pneumatic-electromagnetic-driven impact device integrated with an embedded accelerometer enables accurate and nondestructive detection of mango firmness. The innovative combination of pneumatic control and electromagnetic drive effectively minimizes the impact of fruit size variations and enhances prediction accuracy, demonstrating the significant potential for real-time fruit firmness sorting applications.

The aim of this study was in-line, rapid, and non-destructive detection for soluble solid content (SSC) in pomelos using visible and near-infrared spectroscopy (Vis-NIRS). However, the large size and thick rind of pomelo affect the stability of spectral acquisition and the biological variabilities affect the robustness of models. Given these issues, in this study, an efficient prototype in-line detection system in transmittance mode was designed and evaluated in comparison with an off-line detection system. Data from the years 2019 and 2020 were used for modeling and the external validation data were obtained by the in-line detection system in 2021. The wavelength selection methods of changeable size moving window (CSMW), random frog (RF), and competitive adaptive reweighted sampling (CARS) were used to improve the prediction accuracy of partial least squares regression (PLSR) models. The best performance of internal prediction was obtained by CARS-PLSR and the determination coefficient of prediction ( $R_p^2$), root mean square error of prediction (RMSEP), and residual predictive deviation (RPD) were 0.958, 0.204%, and 4.821, respectively. However, all models obtained large prediction biases in external validation. The latent variable updating (LVU) method was proposed to update models and improve the performance in external validation. Ten samples from the external validation set were randomly selected to update the models. Compared with the recalibration method, LVU could effectively modify the original models which matched the SSC range of the external validation set. The CSMW-PLSR models were more robust in external validations. The off-line model with LVU performed best with a root mean square error of validation (RMSEV) of 0.599% and the in-line model with recalibration obtained RMSEV of 0.864%. These results demonstrated the application potential of the transmittance Vis-NIRS for in-line rapid prediction of SSC in pomelos and the modeling and updating methods could be applied to samples with biological variabilities.

In recent years, worldwide research on fruit and vegetable quality detection technology includes machine vision, spectroscopy, acoustic vibration, tactile sensors, etc. These technologies have also been gradually applied to fruit and vegetable grading and sorting lines in recent years, greatly improving the income of farmers. There have been numerous reviews of these techniques. Most of the published research on fruit and vegetable quality detection technology is still carried out in the laboratory. The emphases have been on quality feature extraction, model establishment and experimental verification. The successful application in the fruit and vegetable sorting production line proves that these studies have high application potential and value, and we look forward to the performance of these sensing technologies in the fruit and vegetable picking field. Therefore, in this paper, based on the future highly automated fruit and vegetable picking mode, we will focus on three kinds of fruit and vegetable quality detection technologies including machine vision, tactile sensor and spectroscopy, to provide some reference for future research. Since there are currently limited cases of detecting quality during the fruit and vegetable picking, experiments performed on prototypes of manipulator, or devices such as Nanocilia sensors, portable spectrometers, etc., which are compact and convenient to mount on manipulator will be reviewed. Several tables and mosaics showing the performance of the three technologies in the detection of fruit and vegetable quality over the past five years have been listed. The performance of each sensing technology was relatively satisfactory in the laboratory in general. However, in the picking scenario, there are still many challenges to be solved. Different from industrial environments, agricultural scenarios are complex and changeable. Fragile and vulnerable agricultural products pose another challenge. The development of portable devices and nanomaterials have become important breakthroughs. Optical and tactile detection methods, as well as the integration of different quality detection methods, are expected to be the trends of research and development.