The potential of Wireless Ad-hoc Control Networks (WACNets) in realising ambient intelligence in a home environment is explored in this paper. The main objective is to achieve energy and resource efficiency while maintaining optimal comfort. WACNets is a novel concept developed by the Centre for Intelligent Mechatronics Research at the University of Wollongong for the purpose of providing a framework for highly distributed, intelligent wireless control networks. A WACNet consists of intelligent nodes developed on the IEEE 1451 Smart Sensor and ZigBee standards. The WACNet platform is fully selforganizing and employs a mesh of star-topology clusters. The development of an intelligent behaviour- learning control algorithm running on WACNet is described in this paper. The learning algorithm is part of an application which aims at reducing the consumption of resources (such as electricity and water). The main objective of this research project is to embed on a WACNet an algorithm capable of learning the factors influencing the use of resources. The WACNet acts as a platform for a highly distributed implementation of agent-based human behaviour learning algorithm, using fuzzy logic. The concept of WACNet is introduced and the test-bed developed for its study is explained. The suitability of WACNets in creating ambient intelligence in a home environment is addressed. A computer simulation developed to demonstrate the concept of fuzzy learning is presented, along with the results of the first test-bed experiments. Finally, some conclusions are drawn.