Design and Implementation of a sustainable Wireless BAN Platform for Remote Monitoring of Workers Health Care in Harsh Environments Karel Heurtefeux, Elyes Ben Hamida, Hamid Menouar Qatar Mobility Innovations Center (QMIC), Qatar Science and Technology Park (QSTP) Po. Box. 210531, Doha, Qatar Emails: {karelh, elyesb, hamidm}@qmic.com Abstract—This paper presents experimental results on a body area network platform that accurately and precisely captures, processes, and wirelessly transmits six-degrees-of-freedom inertial and electrocardiogram data in a wearable, non-invasive form factor. The platform is designed to be low-energy enabling health care applications and remote monitoring of workers in harsh environments. The challenges tackled in this article include the following: (1) reducing the radio channel contention, (2) reducing the energy consumption, and (3) managing diverse Quality of Service (QoS). The system is evaluated regarding to the accuracy and the energy consumption efficiency. I. I NTRODUCTION Wireless Body Area Networks (BAN) are formed by several low-energy wirelessly interconnected biomedical or inertial sensor devices. The sensors may capture various phys- iological parameters of the human body (e.g. temperature, heart rate, Electroencephalography (EEG), Electrocardiogra- phy (ECG), blood pressure, blood oxygen saturation levels, etc.) as well as parameters of the physical environment, such as the amount of sunlight exposure or ambient air quality. In a typical BAN architecture, sensor data are transmitted wirelessly to a gateway where the data are forwarded over Internet to a remote medical server for storage and analysis. Due to constraints such as energy and computation capability, nondeterministic sensor failures, radio links instability, and distrusted environments, designing and deploying a robust BAN platform is still challenging. Motivation. The BANs have become a leading approach for several promising applications in the medical and health- care field. But despite the rich availability of works, there are not many fully functional applications that can be actually employed in real cases. In particular, limited resources in energy and in radio communications make difficult real-world deployment. As a result, there is a need for low-energy communication protocols reducing both traffic contention and energy consumption. In this paper, we present a BAN platform monitoring workers who are subjected to hard environmen- tal conditions during their work. The platform implements similarity filtering and polynomial interpolation techniques. These lightweight mechanisms are suitable for low-power microcontroller and allow to efficiently reduce the amount of data that must be transmitted. In addition, we show that, even with an important compression with loss, they still allow to detect fall or anomaly in heart beat. Contributions. In this article, we address these challenges and make the following contributions: • Firstly, the architecture of the BAN platform is pre- sented and application scenarios are described. Three types of signals are considered: electrical activity of the heart or electrocardiography (ECG), orientation measurement via a tri-axial gyroscope, and linear ac- celeration measurement via a tri-axial accelerometer. • Secondly, similarity filtering and polynomial regres- sion are proposed to provide large compression while maintaining high accuracy. • Finally, the paper presents a scenario allowing to evaluate the performances of the proposed architecture in terms of accuracy, efficiency and energy saving. Novelty. WBANs area has been recently the subject of intense research by many researchers worldwide and there are available many good results in all such topics in particular in efficient physical layer and networking protocols. However, only a few studies related to the development of practical, efficient and low-energy WBANs system have been proposed. In this paper, we experimentally evaluate the accuracy and efficiency of two mechanisms allowing energy saving while staying accurate according to the state of the monitored sub- ject. The rest of this paper is organized as follows. Section II provides an overview of previous work insisting on information which is relevant to the context of this paper: previous experi- mental BAN architectures and existing standards are presented. In Section III, an overview of the proposed BAN architecture is proposed: application scenarios and communication archi- tecture are detailed. In Section IV, the experimental results on the platform and performance are presented. Finally, Section V concludes the contributions of this paper and discusses potential further work directions. II. RELATED WORK AND SCOPE Recent developments in electronics and ultra low power radio communications have enabled the design of tiny and smart wearable sensors which can be worn on, or implanted