Performance and availability evaluation of an smart hospital architecture

Abstract

Low latency and high availability of resources are essential characteristics to guarantee the quality of services in health systems. Hospital systems must be efficient to prevent loss of human life. Smart hospitals promise a health revolution by capturing and transmitting patient data to doctors in real-time via a wireless sensor network. However, there is a significant difficulty in assessing the performance and availability of such systems in real contexts due to failures not being tolerated and high implementation costs. This paper adopts analytical models to assess the performance and availability of intelligent hospital systems without having to invest in real equipment beforehand. Two Stochastic Petri Net models were proposed to represent intelligent hospital architectures. One model is used to assess performance, and another to assess availability. The models are pretty parametric, making it possible to calibrate the resources, service times, times between failures, and times between repairs. The availability model, for example, allows you to define 48 parameters, allowing you to evaluate a large number of scenarios. The analysis showed that the arrival rate in the performance model is an impacting parameter. It was possible to observe the close relationship between MRT, resource utilization, and discard rate in different scenarios, especially for high arrival rates. Three scenarios were explored considering the second model. The highest availability results were observed in scenario A, composed of server redundancy (local and remote). Such scenario—with redundancy—presented an availability of 99.9199%, that is, 7.01 h/year of inactivity. In addition, this work presents a sensitivity analysis that identifies the most critical components of the architecture. Therefore, this work can help hospital system administrators plan more optimized architectures according to their needs.

Model-Driven Impact Quantification of Energy Resource Redundancy and Server Rejuvenation on the Dependability of Medical Sensor Networks in Smart Hospitals

“…The results show optimal settings for system performance and availability. The article by [18] is one of two works that this work extends. The study by Araujo(b) et al [11] proposes an energy availability model for intelligent construction.…”

“…Rodrigues et al [18] propose models capable of calculating performance and availability metrics in a smart hospital system. The work presents a performance model capable of calculating Mean Response Time, Resource Utilization, and Discard.…”

“…The values used were taken from other validated studies. The values were taken from four papers [11,18,38,39]. The TIME_PEAK and TIME_OFFPEAK values represent daytime and nighttime, respectively.…”

“…SPN can represent synchronization, sequencing, parallelization, and concurrency, among other aspects of any distributed system. SPN has already been adopted in previous studies in the hospital context, however, without focusing on energy issues and software aging [17,18]. Therefore, the main contributions of this paper are:…”

Implementasi Raspberry Pi untuk membantu Klinik dan Pengembangan Website Klinik menggunakan Rapid Application Development

“…Pada penelitian [9], yang berfokus pada sistem di rumah sakit dituntut efisien agar tidak terjadi nyawa yang hilang. Konsep pada arsitektur rumah sakit pintar memperlihatkan perubahan kesehatan untuk mengirimkan dan menangkap data pasien yang akan dikirim ke dokter secara waktu nyata memanfaatkan jaringan sensor nirkabel.…”