Abstract
This thesis explores the application of Electrical Impedance Spectroscopy (EIS) as a non-destructive technique for assessing fruit quality, with a focus on detecting physiological changes associated with ageing, mechanical damage, chilling injury, and ripening. In response to the demand for reliable, real-time, and scalable methods in food quality control, EIS offers a promising solution by probing the electrical properties of biological tissues. The research is structured around three main experimental components: contact-based EIS for laboratory-based evaluation, non-contact EIS for industrial scalability, and a combined analysis to assess their complementary strengths. This framework addresses key challenges posed by the heterogeneous structure of fruit tissues, including variations in water content, cellular integrity, and biochemical composition. The contact-based EIS experiments focused on nectarines, examining tissue degradation under cold storage and the progression of mechanical damage following impact. These studies demonstrated the technique’s sensitivity to subtle structural changes, especially when analyzed using equivalent circuit modeling and machine learning techniques. The non-contact-based EIS setup, employing a capacitive configuration, was applied to study banana ageing and mechanical damage in nectarines. These investigations revealed the potential of non-contact EIS for rapid, high-throughput applications while also highlighting limitations such as environmental sensitivity and dielectric inhomogeneity. Finally, the integration of both methods was explored through a case study on avocado ripening, using dry matter content as a reference quality attribute. By applying both contact and non-contact systems on the same fruit samples, the analysis evaluated each method’s effectiveness in capturing physiological trends and demonstrated their combined utility in assessing complex fruit structures. Key findings highlight the frequency-dependent nature of EIS responses and their correlation with fruit tissue condition. A broad frequency range was used to examine impedance and phase spectra, enabling the extraction of physiologically meaningful parameters across diverse fruit types and treatment conditions. Data analysis strategies were tailored to each experiment, ranging from Bode plots and statistical evaluations to advanced machine learning classification, depending on the system’s stability and the experimental objectives. These approaches improved the interpretability and robustness of the bioimpedance data. In conclusion, this thesis establishes a comprehensive EIS-based framework for non-destructive fruit quality monitoring, integrating both precision-focused laboratory techniques and scalable, non-contact industrial methods. The dual strategy enhances adaptability across various applications, contributing to the development of sustainable, efficient, and non-invasive technologies for postharvest quality control of fruit, with a particular emphasis on nectarines, bananas, and avocados.