View-based approach to constructing reliable Home Appliance Control System

Often times we may find ourselves in a situation where we are miles away and recall that we haven't closed our garage door or haven't switched off the stove, and using our internet-enabled wireless mobile device we send a command to our Home Appliance Control System (HACS) to either close the garage door or switch the stove off. However, how can we be sure that the command was executed and that the desired situation, for example, the closed garage door or the switched-off stove, was reached? This paper proposes a technique for constructing reliable HACS's (or RHACS) using the concept of forward view and reverse view, where a view includes not only the physical path but also the control intelligence for that path. RHACS will not only help people remotely control devices at home but also increase confidence that their commands were effective with no unanticipated side-effects. A brief definition of reliability is the probability of good working of a system - however, our survey of the literature indicated that there is no consensus on this definition. Our analysis of a typical HACS indicated that its reliability depended on three major factors - reliability of software, reliability of hardware, and reliability of the network. As a case-study we considered a HACS configuration that included a washer, dryer, garage door opener, stove and a camera, and explored how the reliability of this system could be improved using the view-based approach. Concentrating on the network reliability aspect we explored three different techniques to improve the overall system reliability: standard protocol (X10) for the forward and reverse views, standard protocol (X10) for the forward view and the wired Ethernet for the reverse view, standard protocol (X10) for the forward view and the wireless Bluetooth for the reverse view. We used the NFR Framework to systematically analyze and evaluate reliability of HACS while at the same time accommodating the varying definitions of reliability, and we validated these evaluations using simulations. While further work needs to be done to determine the effectiveness of this approach to other reliability factors, we believe that this study demonstrates the practicality of the view-based approach to methodically analyze and construct reliable HACS with almost negligible overhead.