Description
Abstract – Intravenous (IV) therapy is ubiquitous in acute care, yet gravity-driven drip administration remains highly
dependent on periodic human observation and manual roller-clamp adjustment. Variations in bag height, venous
backpressure, patient movement, and tubing kinks can cause unintended under- or over-infusion between rounds, while
documentation and alarm escalation are inconsistent. These factors increase the likelihood of dosing error and place
additional cognitive load on nursing staff. This paper presents an AI-based smart IV drip controller with Internet of
Things (IoT) connectivity that estimates delivered flow in real time, adaptively actuates the clamp using a low-cost
motorized mechanism, and enforces safety guardrails to prevent overdose and detect abnormal conditions such as
occlusion or line disconnection. The proposed approach combines lightweight regression control with rule-based
safety constraints and human-in-the-loop override. A benchtop-oriented evaluation is described, together with closed
loop simulations demonstrating improved flow-rate tracking and timely anomaly alerts compared with intermittent
manual adjustment. The architecture is designed to support edge operation and secure telemetry to clinical dashboards,
reducing human error while maintaining clinical accountability.
Reviews
There are no reviews yet.