System-level approach to the design of ambient intelligence systems based on wireless sensor and actuator networks

Abstract

Wireless sensor and actuator networks (WSANs) have become pervasive and are used in many embedded and intelligent systems. However, the complexity of applications based on these networks is limited due to lack of tools for designing distributed systems on top of WSANs. In this paper, we present how a system-level programming language, SystemJ, is used to develop a middleware-free Ambient Intelligence (AmI) system. The system consists of a combination of Internet-enabled stationary and mobile WSAN nodes, which resembles an Internet of Things scenario. A distributed warehouse monitoring and control scenario with collaborating stationary and mobile WSAN nodes is used as a motivating example designed and implemented in SystemJ. This example demonstrates the capabilities of SystemJ for designing distributed AmI systems with inherent support for reactivity and composition of concurrent behaviors based on a formal model of computation, without the need for any additional middleware. The approach is compared with existing software agent, robotic and WSAN middleware approaches in designing the same type of systems.