Design and feasibility analysis of a graded harvesting end-effector with soluble solid content estimation function

In response to the prevailing scarcity of labor and with the aim of augmenting the proportion of premium-quality fruits, a robotic grading end-effector system for harvesting has been meticulously designed. The end-effector could measure the soluble solid content (SSC) of peaches during the harvesting process to evaluate the quality of the fruit, thereby facilitating real-time grading during harvesting. Comprising a harvesting component and an information-gathering segment, the end-effector system optimized the primary structural parameters of its adaptive fingers using a mathematical model of peach morphology. Concurrently, the introduction of buffering materials mitigated the pressure exerted by the adaptive fingers on the peaches. Furthermore, feasibility analyses of the grasping actions were conducted based on the interaction forces between the adaptive fingers and the peaches. To grade the quality of peaches, SSC was used as an indicator to assess and grade the quality of the peaches. The spectra of peaches within the wavelength range of 590-1100 nm were collected, and a predictive model for SSC was developed. The correlation coefficients for the calibration set and prediction sets of the predictive model were 0.880 and 0.890, with corresponding root mean square errors of 0.370 and 0.357% Brix, respectively. In addition, a robustness and accuracy assessment was conducted using 30 peach samples, yielding a correlation coefficient of 0.936 and a standard error of 0.386% Brix between the predicted and measured values of SSC. The results confirm that the end-effector can measure the SSC of peaches during the collection process, providing novel concepts and theoretical foundations for real-time harvesting and grading.