The growth of the world's population and the dramatic increase in average life expectancy raise awareness of the global burden of elderly diseases. Mild cognitive impairments affect elderly people, which suffer a deterioration of their capabilities, with special regard to memory. Such problems often lead to disorientation or wandering: erratic movements and changes in mobility and behavioural patterns. In this context, an early and automatic detection of wandering symptoms avoids risks and possible damages to these patients, improving their quality of life. However, wandering detection is not trivial, because several factors such as patterns, environment or the severity of the dementia need to be considered. The research community has proposed several techniques, algorithms and technologies in order to properly detect such abnormal behaviours. Moreover, the new health paradigm, Smart Health, opens the door to a new generation of systems and technologies in order to overcome the shortcomings of the classical methods in this field of research. This paper analyses the most relevant wandering detection methods present in the state-of-the-art.