As the name suggests, Context-Aware Computing (CAC) consists of computational systems that are context-sensitive. That is, collecting data from the environment in which they operate enables them to interpret their context. In this case, the word context refers to “information which contains a detailed description of object features or environment state”. This information is used to fully describe the object, which can be in terms of features such as functionality or method of operation (Schmidt, 2005;
Fischer et al., 2004; Abowd et al., 1999).
Due to the AmI complexity, context is composed of different things (e.g. devices, environmental conditions, humans), each one with its individuality. Applied to humans, context is defined as the performed actions, location, interaction with other people and social status. In computing science, context can be defined as available resources, connectivity, bandwidth and results, among others. In home, context can be the features of an appliance or other devices, and their current state. So,
depending on the area that is observed, the context includes different elements and has different structure.
The CAC mainly hosts developing applications that use context characterizing information to optimize the functions to be performed. In practical terms, it results in two types of applications: the first ones directly operate on the environment according to the context changes, while the others act upon other applications feeding them with information about the context. There are numerous applications developed in each group. So, for example, applications whose performance changes when the user has a disability or even interfaces tailor to the user. Therefore, these applications are embroiled with the user profile context, using information such as age, eyesight and hearing or memory capacity.
From that, it can be concluded that the crucial issue is the context setting as well as the required contextual information for a given object. Additionally, other key issues are the connection between different objects, what properties they share and how they complement each other. Subsequently, the information must be classified to result in a format suitable for the system.
Thus, the aim is designing a system able to adapt to environment changes by considering user preferences and needs. To achieve that, sensors and user profile are used since they respectively provide information about the context environment (e.g. position of the wearer, room temperature, or luminosity) and user context (e.g. health conditions such as heartbeat). In that way, a system is designed such that it is in accordance with the response to be performed, while considering all involved contexts and the intended goal.
The context concept can be shaped to gather different types of information coming from static sensors, body sensors or the user's calendar. Given the flexibility concept and its ability to adapt, it is not difficult to realize its importance. Currently, the sensor fusion data is used to take advantage from the complete information. The sensor fusion concept consists of merging information of different sensors, that have little in common, but their data complement each other. So, for instance, joining data from luminosity, temperature and humidity sensors can result in information about the current and future weather and also lead to home climate control (Sanfeliu et al., 2010; DeLong, 2003; Ercan & Erdem, 2011).
All these differences arise the question what the definitions of context-aware applications are. The definition of context demands a rule to create applications. As incoming data is variable, the application is, too. However, ontologies serve as communication normative and can ensure interoperability between other systems platforms. In terms of the used devices, the system must have a detailed
characterization of them and be able to share this information with the other platform services. Thus, it is easier for the services to adapt to the devices and advantage from its features.
The CAC is useful for AmI because it defines the basic concept of intelligent actions and allows the use of devices with actions considered intelligent. These devices respond to the change of states, captured by sensors, ruled by the context in which they operate and how they are programmed.
Context-aware computing helping disabled people
The development of systems that help disabled people is being discussed by the European Community, resulting in several alternatives for such problems. Some alternatives are more complex and expensive than others, but undoubtedly the technological solutions are a viable alternative.
The CAC, with its features, is perfect for applying adaptive systems that respond to the users and their needs. The elderly are more susceptible to health problems, suffering from several constraints that limit their lives, being the most common limited motor skills, hearing or vision problems and cognitive impairment. The CAC can conveniently provide answers to specific user needs, taking into account user health problems, the environment and the objects that make it up. This can be achieved by means of a carefully built context concept, selecting what is really needed and what can be ignored. The features which should have access to these systems are:
− Emergency assistance: detects and accordingly acts when critical events occur;
− Greater autonomy: through domotic systems and applications that liberate the user from complex and arduous tasks;
− Increased comfort: providing safety, maintenance and house cleaning, as well as managing personal events and leisure time.
These systems keep a constant assistance to users. The scientific community has coined the ecosystems that contain CAC system working in favor of the user as Ambient Assisted Living.