Dia-Smart is an IoT-based diabetes-management ecosystem that connects insulin-dose tracking, glucose readings, storage safety, inventory monitoring, alerts, and caregiver visibility through one integrated platform.
Elderly insulin users must coordinate glucose measurement, insulin administration, safe storage, and inventory management. Existing devices usually operate independently, creating incomplete records and increasing dependence on memory and manual tracking.
Missed, repeated, or incorrectly recorded insulin doses.
Insulin may be exposed to unsuitable refrigerator temperatures.
Patients may not identify low insulin availability in time.
Caregivers and doctors lack a single, complete patient timeline.
Dia-Smart combines three connected hardware units with a secure cloud platform and role-based dashboard to create a unified patient-management workflow.
Installed inside the refrigerator to monitor insulin-storage temperature, remaining inventory, and storage activity.
Acts as the elderly-friendly patient interface, local processing unit, fallback confirmation device, and cloud communication gateway.
Attaches to the insulin injector pen and uses a magnetic encoder to detect dial rotation and estimate the selected insulin dose.
Combines dose, glucose, storage, inventory, prescription, and alert records within one patient timeline.
Dia-Smart brings together physical hardware, wireless networks, and secure software applications to streamline daily clinical operations.
Uses magnetic-encoder readings to monitor insulin-pen dial movement and record estimated dose events.
Retrieves readings from supported Bluetooth Low Energy glucometers.
Tracks refrigerator temperature and storage activity.
Uses weight measurements to estimate the remaining insulin quantity.
Provides relevant system notifications and remote access to patient records.
Buffers important events locally and synchronizes them after reconnection.
Allows the patient to confirm or correct uncertain dose readings using the Outer Unit display and keypad.
A four-layer IoT architecture connecting sensing devices, secure communication, cloud processing, and role-based user interfaces.
Custom microcontrollers, sensors, and structural enclosures integrated to support daily routines.
Firmware has been developed for the Inner Unit, Pen Unit, and Outer Unit using the ESP32 Arduino framework and PlatformIO. Modular services and FreeRTOS-based tasks allow sensing, communication, storage, and user-interface operations to run independently.
The software tier coordinates clinical tasks, aggregates records, and manages user views securely.
The Spring Boot backend processes MQTT telemetry, applies business rules, manages users and devices, generates alerts, and provides secure REST APIs.
A responsive React web interface accessible on desktop, tablet, and smartphone web browsers.
Implemented on AWS cloud infrastructure for scalability and security.
Continuous integration checks run alongside hardware validation benches to confirm system functionality.
Consolidated prototype construction costs and testing expenditures.
Schedule of the research, development, and validation milestones spanning November 2025 to August 2026.
Dia-Smart currently has a working basic integrated prototype. Current testing focuses on sensor and dose-detection accuracy, device communication reliability, offline buffering, cloud synchronization, alert generation, dashboard usability, long-duration operation, and enclosure refinement.
The current Dia-Smart Pen Unit already supports dial-rotation sensing, dose estimation, embedded processing, and wireless transmission through a compact attachment fitted to a HumaPen Ergo injector pen. The future plan is not to add a missing feature, but to redesign the validated mechanism as a complete Dia-Smart insulin pen with an integrated smart cap. This will replace the external commercial pen body with a more compact, adaptable, manufacturable, and fully proprietary Dia-Smart design.
Dia-Smart sensing and communication attachment fitted to a commercial HumaPen Ergo injector pen.
Purpose-built Dia-Smart insulin pen and integrated smart cap retaining the validated sensing and communication functionality.
Department of Computer Engineering, Faculty of Engineering, University of Peradeniya.
Visual log of hardware component manufacturing, laboratory validation tests, and interfaces.
Reference sources, code repositories, and institutional websites.