An event-based sampling method is developed for wireless smart appliance networks. This is accomplished with a Kalman filter that includes an asynchronous measurement update along with a standard model-based prediction of the state- variables. This approach minimizes the number of communication transmissions while also providing rapid updates on an as-needed basis to those appliances that require attention. This work is motivated by recent demands for household energy management systems that coordinate in real-time the optimal allocation of electrical usage among various household appliances. Results for a household network are provided for the case of a networked variable speed clothes dryer.