Smartphone-based Sensor System for Real-Time Food Spoilage Detection

Thesis Abstract

Food spoilage remains a critical challenge in ensuring food safety, reducing wastage, and maintaining nutritional quality, particularly in developing countries where traditional inspection methods are often inadequate or inefficient. The rapid detection of food spoilage at the point of purchase or storage can significantly mitigate health risks and economic losses. This study aims to develop and evaluate a smartphone-based sensor system capable of real-time detection of food spoilage through passive and active sensing technologies integrated with mobile computing. The specific objectives are to design a cost-effective sensor interface compatible with smartphones, establish reliable detection protocols for common spoilage indicators such as volatile organic compounds (VOCs), pH changes, and microbial activity, and validate the system's performance across varied food matrices including dairy, poultry, and fresh produce. The research employs a mixed-methods approach, combining experimental laboratory tests with a field evaluation in a real-world setting. The experimental phase involves collecting data from 150 food samples (50 each of dairy, poultry, and produce) subjected to controlled spoilage conditions, with sensors measuring key chemical and biological parameters. The sensor modules incorporate gas sensors (e.g., metal-oxide sensors for VOC detection), pH microelectrodes, and microbial growth indicators, all interfaced with a custom-designed mobile application to facilitate real-time data transmission and analysis. The field evaluation involves 100 consumers and vendors using the prototype system over a three-month period to assess usability, reliability, and scalability. Data analysis utilizes quantitative methods including multiple regression analysis to determine the relationship between sensor outputs and microbial load as measured by standard laboratory techniques (plate counts, PCR). Receiver operating characteristic (ROC) analysis assesses the system's sensitivity and specificity in spoilage detection. Thematic analysis is applied to qualitative feedback obtained from users regarding system usability and potential integration into existing food supply chains. Expected findings suggest that the sensor system provides accurate, rapid, and non-destructive assessment of food freshness, demonstrating high correlation (r > 0.85) with laboratory microbial analyses. The system is anticipated to achieve sensitivity and specificity rates exceeding 90% in identifying spoiled samples across different food categories. The integration with smartphone technology is expected to facilitate user-friendly interfaces and data sharing capabilities, promoting widespread adoption at various points in the food supply chain. This research advances current knowledge by demonstrating the feasibility of low-cost, portable sensor systems integrated with mobile devices for food spoilage detection. It contributes to the theoretical framework of food safety monitoring by extending the application of biosensors and ICT tools, grounded in the Technology Acceptance Model (TAM) and diffusion of innovations theory. The study also offers empirical evidence on the accuracy, reliability, and user acceptance of smartphone-based sensor systems in food safety contexts, addressing existing gaps in real-time, non-invasive spoilage detection technologies. The main conclusion emphasizes that smartphone-based sensors can effectively complement or potentially replace traditional inspection methods, offering a scalable, accessible solution for improving food safety and reducing waste. Recommendations include further refinement of sensor sensitivity, integration with supply chain management systems, and regulatory validation. Future research should explore the deployment of this technology across diverse geographic and socio-economic environments to enhance global food safety monitoring and management practices.