RFID based Health Adherence Medicine Case Using Fair Federated Learning

TL;DR

The paper tackles medication nonadherence and privacy concerns in health monitoring by proposing a Smart Pill Case that fuses RFID-based data recording, NFC data extraction, a load cell for precise dosage measurement, and an Android app for real-time feedback. It grounds the work in fair federated learning to enable cross-user learning while preserving privacy, and reviews existing adherence tools to motivate an IoT-based, patient-centric solution. The design combines a hardware stack (load cell, RFID, Arduino) with NFC-enabled data storage and an accompanying mobile interface, aiming for accurate event detection, timely warnings, and scalable data sharing. Experimental results on unit-weight calibration and the end-to-end hardware-software workflow demonstrate feasibility and highlight opportunities for improving battery life, multi-medication support, and cloud-based collaboration with clinicians and families.

Abstract

Medication nonadherence significantly reduces the effectiveness of therapies, yet it remains prevalent among patients. Nonadherence has been linked to adverse outcomes, including increased risks of mortality and hospitalization. Although various methods exist to help patients track medication schedules, such as the Intelligent Drug Administration System (IDAS) and Smart Blister, these tools often face challenges that hinder their commercial viability. Building on the principles of dosage measurement and information communication in IoT, we introduce the Smart Pill Case a smart health adherence tool that leverages RFID-based data recording and NFC-based data extraction. This system incorporates a load cell for precise dosage measurement and features an Android app to monitor medication intake, offer suggestions, and issue warnings. To enhance the effectiveness and personalization of the Smart Pill Case, we propose integrating federated learning into the system. Federated learning allows the Smart Pill Case to learn from medication adherence patterns across multiple users without compromising individual privacy. By training machine learning models on decentralized data collected from various Smart Pill Cases, the system can continuously improve its recommendations and warnings, adapting to the diverse needs and behaviors of users. This approach not only enhances the tools ability to support medication adherence but also ensures that sensitive user data remains secure and private.

Figures (15)

• Figure 1: MEMS Cap (the top part) will detect the date and time when the bottle is opened over 3s, the pill bottle can be any bottle as long as it fit with the cap [9].
• Figure 2: Helping Hand (the blister sleeve) will detect the date and time when the blister is removed from the sleeve and reinserted, the blister pack must be in a particular dimension, the circular LCD will display the dosage within the last 24 hours and the time since the last dose [9].
• Figure 3: Smart Blister equip with self-adhesive label (the block at the left of the blister) will detect the date and time when a pill is removed from the blister, the blister pack can be any blister pack available [9].
• Figure 4: The overall information flow of the system, the action refers to the action of taking pills out of the case, the upper hardware part occurs in the Smart Pill Case and the lower software part occurs in the Smart Phone.
• Figure 5: The wiring diagram of the Smart Pill Case, the load cell is a 1KG portable load cell, the HX711 is the load cell amplifier, the RFID-RC522 is a 13.56MHz RFID module, they are all connected and controlled by Arduino Uno.