Background and Motivation
Intravenous fluid administration is one of the most routine and critical procedures in healthcare. Nurses and doctors must monitor IV bag levels continuously to ensure patients receive the correct fluid volume and to prevent the line from running dry, an event that can introduce air into the bloodstream, cause electrolyte imbalance, or in severe cases trigger cardiac events. The global error rate associated with IV medication administration is estimated between 9.4% and 97.7%, and approximately 15 to 20% of patients receiving IV treatment in hospitals experience complications linked to inadequate monitoring. In busy clinical settings, continuous manual observation of IV bags is neither practical nor reliable.
Proposed System
This paper presents a prototype IV Bag Monitoring and Alert System that automates fluid level detection and delivers real-time alerts to healthcare professionals without requiring any contact with the bag or its contents.
The core sensing mechanism is a non-contact capacitive fluid level sensor mounted externally on the IV bag via a Velcro band. Capacitive sensors detect changes in the dielectric properties of the surrounding medium, in this case the presence or absence of fluid behind the bag wall, allowing accurate level measurement without puncturing, touching, or contaminating the bag in any way. This non-contact design makes the device compatible with any standard IV container geometry and eliminates the risk of sensor-introduced contamination.
The sensor is connected to an ESP32 microcontroller, which continuously reads the fluid level, processes the signal, and triggers alerts when the level drops below a predefined threshold. The alert mechanism operates on two channels: a local audible beep signal at the bedside, and a push notification delivered to healthcare staff through the Blynk platform, a cloud-connected IoT application accessible via mobile or web interface. A centralised web-based dashboard allows nursing staff to monitor multiple patients' IV bags simultaneously from a single station, enabling remote surveillance without physical presence at each bedside.
Results and Significance
The prototype successfully detects low fluid levels in real time and delivers alerts through both local and remote channels. The system is compact, low-cost, and non-invasive, requiring no modification to the IV bag or administration line. Its compatibility with various container types and its wireless connectivity make it suitable for deployment across diverse clinical environments including general wards, ICUs, and outpatient infusion centres.
The primary contribution is the demonstration that a reliable, low-cost, non-contact IoT-based solution can replace continuous manual monitoring for IV fluid management. By automating the detection and alert process, the system reduces the burden on nursing staff, minimises the risk of human error, and improves patient safety outcomes. The integration with a centralised platform also enables scalable multi-patient monitoring, which is particularly valuable in high-volume or understaffed clinical settings.
Significance in the Broader Research Portfolio
This work represents the earliest published contribution in this research programme and establishes the foundational motivation that runs through all subsequent work: the application of technology to address gaps in clinical monitoring and patient safety. While the later papers address software-level failures in clinical AI, how models are evaluated, calibrated, and deployed, this paper addresses the hardware layer, demonstrating that patient safety improvements can be achieved through thoughtful sensor design and real-time alerting, even with lightweight and inexpensive components.