A Developed Design of Low-Cost Insulin Pump for Automated Insulin Delivery (Aid) System

Abstract

Diabetes management requires precise insulin delivery to mitigate life-threatening risks from glucose fluctuations. Current systems face challenges in cost, reliability, and integration with continuous glucose monitoring (CGM). This study introduces the MARS-Z infusion pump, a low-cost, compact device integrated with CGM for real-time glucose tracking and monitoring via mobile application. The system employs advanced sensor fusion (encoder, potentiometer, current sensor) enabling high-precision delivery verified by redundant position sensing (99% dosing accuracy in preliminary tests) and proactive fault detection via real-time electrical signature monitoring. While the current prototype relies on manual dose input, its architecture is designed for future integration with dosing algorithms. Wireless connectivity allows remote monitoring via mobile app, and the prototype's cost represents a 90% reduction compared to commercial pumps. Future work will focus on implementing AI-driven algorithms for automated dose calculation, addressing dynamic responsiveness. This innovation lays the foundation for a fully automated system, prioritizing affordability, safety, and scalability in diabetes care.