A Scalable Multimodal Data Acquisition Infrastructure for Passive Dietary Monitoring Using Consumer Smartwatches and Mobile Technology

raditional dietary assessment methods, including food frequency questionnaires and 24-hour recalls, are subject to substantial recall bias and participant burden. Wearable sensor-based approaches offer the potential for passive detection of eating episodes, but their translation into scalable research protocols remains challenging. We describe the design and deployment of a multimodal data acquisition infrastructure developed within the FITMATE pilot (DARE/PNRR program) to enable continuous, passive dietary monitoring using consumer-grade smartwatches integrated with electronic data capture and centralized data management.