Smart IoTs based urine measurement system.

Urine Measurement is one of the most important processes for diagnosis in the hospitals nowadays. Acute Kidney Injury (AKI) is usually diagnosed by taking patient’s urine samples for a specific period of time. It has been suggested that the average Urine Output of a patient depends upon his weight. As we are all aware that currently the means to monitor the major vital signs of the human body in the ICU (Intensive Care Unit) or various clinical settings such as Heart Rate, Blood Pressure, Central Pressure etc. is done by the means of a continuous recording of impulses and its digital display. It is utmost necessary to record and continuously monitor a patients’ fluid input, administered mostly by electronic devices (e.g. Syringe infusion pumps). At the same time, it is also important to monitor patients’ fluid outputs, in which, urine volume is one of the major components. Currently, it is obtained intermittently (per hour) from urine meters and urine collection bags, and a visual assessment is made and recorded manually relying heavily on the nurse's capability and skills. Therefore, even after so much technological advancement the measurement of urine output is literally the only critical parameter constantly recorded and monitored non-electronically by the medical staff. The references from Medical Professionals at Royal Bournemouth Hospital clearly indicate a need for automated Urine Measurement System for efficient diagnosis process. There are automated devices for urine measurement, that are discussed in the Literature Review section, but none of them is available commercially. Some have cost issues whereas others are too complex to implement. We have found approx. 15 systems which have been patented by the inventors but none of them made it to the market. Cost-efficiency, complexity, and reliability are the issues we need to address, and we have tried to address in our project. In this project, an integrated prototype based on IoT, that measures urine volume in real time for both high and low flow is developed. The system measures the urine coming from the patient through two different sensors, Photo Interrupter Module and Hall-Effect based liquid sensor, and transmits that data to a cloud-based application via WiFi. The Arduino Yun micro-controller was used because of its built-in WiFi chip and more robust performance as compared to other options. The measurement of both high and low flow of liquid makes our system unique from the existing systems. The application at Cloud analyze the data from the sensors for visualizations as mentioned by the doctors. MATLAB analytics facilities will be used because it provides extended options for multiple real-time visualizations. The data is sent in real-time, every 20 seconds and visualizations are updated accordingly. The data is also available to view on an Android App. The real-time stream of data on cloud and ease of data accessibility distinguishes our system to those described in the literature. Series of experimentation was carried out for the prototype. Firstly, due to a problem in Photo Interrupter sensor for drop by drop measurement, the error was huge. Then, we developed an algorithm that solved the problem of object detection and then the error came to below 10% for both the high and low-flow measurement combined. This algorithm can be used to improve the working of photo interrupter sensor in other scenarios and it is one of the contributions of our project. This system decreases the workload of the nursing staff as well as that of the doctors. The human-error is minimized. The Data Analytics application enables the doctors to have an in-depth understanding of the condition of a patient at several different intervals of time. Hence, our system is expected to benefit the medical industry and especially the staff at the hospitals. Lastly, we have also found our concept to be helpful in process industry also where the liquid measurement is used and we presented this concept at EPSRC conference in Glasgow