Background - Smartphone-Based Human Activity Recognition - page 8

when advances in network technologies (e.g. internet, wireless communications)
were incorporated in the original idea (Abowd andMynatt 2000 ; Cook andDas 2012 ).
The integration of new technologies in our everyday life is evolving very rapidly.
The miniaturization, greater functionality, and lower manufacture costs of electronic
devices and sensors are making a transformation in the way humans and machines
interact. These technological changes are allowing the development of AmI services
to seamlessly face user needs (Cook andDas 2012 ).Moreover, dealingwith a growing
population, the lack of sufficient human resources in some areas (e.g. caretakers,
security forces, medical staff), and limited budgets are also important aspects which
encourage people, communities and countries tomake use of AmI for their advantage.
AmI can benefit all population sectors. However, some of them, such as people
with limitations (e.g. the disabled and the elderly), have become a priority and need
immediate attention. In consequence, there is an area of AmI which is specifically
focused on the development of intelligent systems to assist these people: Ambient
Assisted Living (Kleinberger et al. 2007 ). This area is developing very rapidly in
accordance to the current population explosion statistics of these age groups and
their growing demands (Aarts and Wichert 2009 ;Eur 2011 ).
In this section we explore the principles and characteristics of these two areas
of study and explore how they interact with each other. Finally, we connect these
concepts with the development of HAR systems.

2.2.1 AmI Systems

AmI is still in its early stages but current technologies are facilitating its development
and the introduction of new systems (Kleinberger et al. 2007 ). These AmI systems
exhibit common characteristics some of which are here described:

• They are unobtrusive : hardly or not perceived by its users. This avoids disturbances

limiting users during the execution of their normal activities. For example, using
smart wearable accessories (e.g. smartphones, watches and glasses) and devices
embedded in clothes, nearby objects and furniture.

• They communicate between nodes. This involves fast and robust communication

mechanisms, such as wireless technologies, to connect different devices (nodes)
when the transmission of information over a distance is required.

• They are ubiquitous : available at any location, at any time. This characteristic con-

siders the continuous delivery of services, through the use of either fixed infrastruc-
ture, mobile equipment or their combination, regardless of possible limitations
linked to time and location.

• They are context-aware . They capture the state of the user and its environment by

exploiting heterogeneous sensors in order to respond accordingly to diverse and
changing conditions.

• They intelligently anticipate to our needs based on previous information (e.g.

activity patterns, past events and their solutions). This is generally done through