A modular wearable device for vital signs monitoring in heavy industries

This thesis deals with the design of a modular wearable prototype able to perform the monitoring of several physiological parameters. Wearable Health Devices (WHD) have drawn a lot of attention during the last decade in a wide range of domains. A variety of commercial devices have been produced with the aim to monitor the health condition of a person in everyday life or during sport activities. The focus of this project is the experimentation of a wearable device intended to be used by workers in heavy industries in order to monitor their vital signs during working days. The main motivation behind this choice is the lack of WHD for this particular type of application. By its nature, the Oil&Gas industry is plenty with hazards to the health of employees. Despite standard and precaution, workers suffer injuries on the job every day. In this respect, WHD can certainly make a valid contribution to improve the workplace safety in dangerous environments. The approach adopted in this work is to implement a single system which included different types of sensors in order to combine all the information extracted, thus providing a complete picture of the health condition of workers. Data measured from each sensor are managed by the Atmega328 microcontroller. A firmware based on the use of Interrupts events is designed to perform the following operations: data aquisition, data processing and data trasmission. A stressfull situation, due to the onset of a illness, triggers several body responses. Changes in the normal heart functioning is the first evidence. For this reason, an electrocardiogram (ECG) to measure Heart Rate (HR) parameter is included. A temperature sensor to monitor the skin temperature is also included in the prototype. In response to events which pertubate the normal state of a person, the sympathetic nervous system actives the sweat secretion causing a phenomen known as Galvanic Skin Response (GSR). The proposed system extracts features from GSR signals in order to recognize the variation of the skin electrical conductance, thus identifying the onset of a stress event. A reflectance pulse oximeter made it possible the measure of two important parameters: Peripheral Oxygen Saturation (SpO2) and Pulse Transit Time (PTT). Regarding the extraction of the first parameter, this work proposes an algorithm which calculates the differents portions of light absorbed by the pulsatile and non-pulsatile arterial blood. Furthermore, calibration and testing procedures are carried out taking as reference a commercial medical pulse oximeter. The PTT measurement is performed combining the information given by the two synchronized ECG and PPG signals. Future works should include the implementation of a PTT-based system to estimate Blood Pressure values allowing a more complete health monitoring. Finally, an accelerometer-based fall detection system is implemented to distinguish the different person’s movements and detect a fall event which can be associated to injuries in hazardous environments. Testing on different subjects are carried out to evaluate the accuracy of the algorithms implemented. In order to obtain a broader picture of the system behaviour, different simulations are taken into account during tests procedures. The good performances obtained by the proposed system provide solid basis for the continuation of the prototyping development.