To improve the living standard of urban communities and to render the healthcare services
sustainable and efficient, e-health system is experiencing a paradigm shift.
Patients with cognitive discrepancies can be monitored and observed through the
analyses of power consumption of home appliances. This paper surveys recent
trends in home-based e-health services using metered energy consumption data.
It also analyses and summarizes the constant impedance, constant current and
constant power (ZIP) approaches for load modelling. The analysis briefly
recaptures both non-intrusive and intrusive techniques. The work reports an
architecture using IoT technologies for the design of a smart-meter, and fog-computing
paradigm for raw processing of energy dataset. Finally, the paper describes the
implementation platform based on GirdLAB-D simulation to construct accurate
models of household appliances and test the machine-learning algorithm for the
detection of abnormal behaviour.