Abstract Ambient Intelligence refers to environments which are sensitive and responsive to the presence of people. Application perceptual control in AmI postulates that a system's behavior is a means of controlling its perceptions and behavior is controlled input. Three main phases of transformation of entrance influences in abstraction of increasing levels are defined. Only on last phase of abstract representation take place logging and acceptance of information.
Keywords Ambient Intelligence, Perceptual control, Embedded Systems, Consciousness.
The Ambient Intelligence (AmI) paradigm builds upon ubiquitous information technology, which will become invisibly embedded in human natural surroundings and present whenever person needs it. AmI profiling practices human-centric computer interaction design and is characterized by systems and technologies that are:
- embedded: many networked devices are integrated into the environment;
- context-sensitive: these devices can recognize peoples and theirs situational context;
- personalized: they can be tailored to peoples needs and attuned to all their senses;
adaptive: they are autonomous and enabled by simple and effortless interactions and can change in response to replacement of person;
anticipatory: they can anticipate person's desires without conscious mediation.
In an AmI world, electronic environment is capable to answer needs, habits, behaviour of people, defining at the same time, which exactly is present at this moment at this environment. AmI devices work in concert to support people in carrying out their everyday life activities in easy, natural way using information and intelligence that is hidden in the network connecting these devices. As these devices grow smaller, more connected and more integrated into existing environment, the technology disappears into surroundings until only the interface remains perceivable by users.
Thus, this paradigm merge the newest network technologies, allowing to create networks from mobile information devices, and on their basis to form the digital environments reacting to preferences of users being in them and working with interfaces of public using.
In process of development AmI, work with the AmI-agents begins to occupy the increasing part of a human life. Development portable and implanted AmI-agents will result to that the majority of people become surrounded by exoframework (EF), carrying out search and processing of the information and save a part of memory of his carrier. Thus, EF realises a cogitative work and actually serves the person-carrier for intelligence amplification.
The added reality, arising in so doing in the process, will erase border between the person interior and the objective reality. The person will perceive a reality already sated additional information prompted AmI-agents.
Interacting with the Ambience in AmI allows adaptation on two different levels: on the one hand different users with different abilities can interact with AmI through a customized configuration of interaction devices supporting their individual needs. On the other hand AmI's output can be adapted to user context within different environments, e.g. in the car, in the living room, in the office and in the works.
AmI is a vision on the future of consumer electronics, telecommunications and computing. The emerging AmI vision has the potential to fundamentally change our world. Now are necessary additional interdisciplinary researches, which are capable to integrate such various disciplines, as designing of the equipment, development of the software and a science about the person. Absence of the general understanding of interaction of subsystems and the difficulties encounter at integration of devices of a miscellaneous level, do system integration by very complex employment.
This article addresses basic aspects of AmI, to support human contacts and accompany an individual's path through the complicated modern world. It investigates and brings to light the challenges in system design towards fulfilling the AmI promise. For this purpose, the analysis of "information phenomenon" is carried out and condition of "information contact" is certain.
The system in AmI, as a result of each act of multimodal interaction with objects of an environment, changes (or not) the condition. Energetically the most economical form of interaction of system with an environment is information contact". Such contact is possible, if in system the subject domain of interaction is a priori certain, models of possible interactions are generated and value of each of them is estimated.
Suppose, that the analyze system is internally susceptible to (k) simple physical substances, i.e. it is capable to get and lose through border with environment these substances. Such system can be defined as possessing (k) internal degrees of freedom. Where internal degrees of freedom are understood as opportunities of interactions with surrounding potentially incorporated in system. The set of sensors (Se) defines a subject domain of interaction with environment for system.
However realization of available opportunities of interaction depends not only on properties of system, but also from properties of an environment, represents by unaffected variables (U.V.). In fact last, on border with system has quite certain degree of freedom (l).
Interaction between environment disturbances and system is possible only on the interfaced degrees of freedom when environment and system are simultaneously capable to give and perceive corresponding substances. Name number degrees of freedom of system interfaced among themselves and environments as (n).
The value (n) characterizes able to be realized actually possible interactions between system and an environment, that is defines external degrees of freedom the systems depending on properties of an environment. Absence of any concrete degree of freedom - internal or external - indicates about internal or external isolation of system in relation to corresponding substance.
The considered conditions of interaction between environment and system have following feature. The number of degrees of freedom of an environment always considerably exceeds number of internal degrees of freedom of any system (l>>k). Thus, the possible behavior of system should include perceptual control by changing isolation for forming comfortably external condition.
In contrast with other theories of behavior, which assume that behavior is a function of perception and that perceptual inputs determine or cause behavior of systems, application perceptual control postulates that a system's behavior is a means of controlling its perceptions and behavior is controlled input. Systems do not produce behaviour by computing output. Instead, they produce behaviour by comparing inputs with desired inputs, and using the difference to drive output. Systems do not "respond to stimuli" rather they control input variables. The reference variable for each negative feedback control loop in a control hierarchy is set from within the system.
Perceptional control architecture show in fig.1 is a model of how a system being must be internally organized to accomplish this process called controlling. Perceptional control focuses on how system looks" at experience things, and the way these perceptions are compared with desirable experiences. The unbalance from comparator ( ) produces action (Ac) proceeding from perceptual signals and intelligence (In). Thus model explains how thoughts" become actions, feelings and results, and its principles can be applied to any activity.
Stated broadly, perceptional control explains how entities control what happens to them. It explains the relationships between actions (A) and goals, perceptions and actions, and perceptions and reality, and it does so within a single, testable concept of how systems work. Systems behaviours are seen as the observable consequences of actions aimed at producing intended inputs, it is goal directed and purposeful, not mechanical and responsive. According to perceptional control, behaviour is the control of perception rather than the response to a stimulus. Behaviour is neither just caused by stimuli in the environment nor is it blind execution of internal plans. Behaviour is not an end result; it is an integral part of the closed loop process which controls perception.
As the concepts of homeostasis and homeorhesus may be applied to internal environments, similar they can also be applied to the external environment of system. In body, the brain is a means toward transferring perceptual signals derived from the external environment into the internal environment of billions of interconnected neurons. Control systems within the brain and body are responsible for keeping perceptual signals within survive able limits, regardless of the nature of the environment that they are derived from. By examine model, the future of systems self-awareness might be realized by individuals willing to do so.
The information phenomenon.
Generally, the system has Embedded Systems (ES) with function of intelligence for maintenance of the ability to activity by control of own sensors and actuators. It accepts input signals of sensors (input functions I.F.) and internal signals of a condition of actuators (output functions O.F.) realizing behaviour (B) systems.
Degree of knowledge of system about regularities, to which it should obey, is proportional to its complexity. On conditions that knowledge of system about quality of the own behaviour, signals from an environment can provide channelling of behaviour of system according to the existing in it program (rules of the functioning). The algorithm of the program provides processing of the accepted signals so that the system could keep the entity.
The environment proves directly for system sensors when enters into its subject domain. Interaction process begins when the changing object of an environment pass to system a part of the self condition in the form of the substance with certain properties. Consequently, the system realize one of the self condition on the basis of "sensitivity" to certain quantity of the involving substance and "adjustment" for corresponding qualities of substance of object of an environment. Change properties of system by properties of the substances received from influencing object characterises data reception by system.
Data reception by system represents transformation of a dynamic signal from sensors to a static code of reflection of external influence. Such coded integrated reflections of an external world create effect abstract environment copies (0I). It is the first phase of abstract representation of influence on system. Hence, the first type formed abstraction, is direct reflection of the phenomena of an environment. With the termination of external influence, fixing of abstraction and the corresponding modified condition in system does not occur. Such effect characterises presence of zero level of complexity at system.
On the second phase of abstract representation of influence on the system, corresponding to it changes of systems internal properties are fixed, and in it formed (f) self-image. The "self-image" is the real object formed as a result of functional, physical irreversible transformation of internal properties of system by properties of external object influencing it as a result of established relationships. The "self-image" is the indicator of presence of the first level of complexity at system (1Y). This condition of fixing represents "knowledge" of system which can be shown as the second type of abstraction (1I).
Process of data acceptance by system is a process of tracing (s) changes of an own condition (Y-1(0)) of zero level, which variations occur as a result of interaction of system to changing surrounding objects. As a result of this tracing created setting self-image which is remembered. To this first level corresponds a system condition (1Y(1)).
At the first level from a condition of system of zero level, is formed subjective self-image current condition and the system passes in a condition of the first level.
It is level of knowledge of system about itself and about environment. At this level the system operates with the knowledge. The knowledge is a specific reflection (3) of environment by system. The knowledge is saved in system and allows to be guided in the past.
On the third phase of abstract representation of influences, on the basis of subjective data it is formed (g) a subjective image of "self-image. The image of "self-image lack status of an external reality for system and represents internal model of its condition and behaviour. The model is the indicator of presence of the second level of complexity at system (2Y).
Formed at this level, as a result of tracing (r) the chain of code combinations, gives effect of an initiation of relationships between self-images, effect of formation of the program of ability to activity of system. On this basis the third type of abstraction is created and there is an information acceptance (2I). Such the abstraction of abstraction grows out of the abstract coding of the allocated properties opening essence of perceived objects of an environment and represents their ideal abstraction which gives a material for comprehension and thinking. So there is a knowledge system of oneself known. At this level the system operates with the conscious knowledge.
Y0(2) - it is result of tracing (r) by system of own "knowledge". Together with it, it is tracing (t) by system of the own behaviour (B), i.e. occurrences in system of knowledge of regularities of the own functioning. Comprehension is a universal reflection of environment by system, is a reflection not only set of external objects, but also reflection of such abstraction as regularities and relations
Properties of a condition (Y0(0)) will be transformed to new "knowledge" (Y0(1)), on which basis the new model (Y0(2)) is under construction. Together with it, the current behaviour of system also is reflected (t) in model. The created model is saved in conscious memory and participates in formation (h) the future at the following (T1) external influence.
On the fig.2 three main phases of transformation of entrance influences in abstraction of increasing levels are defined. Only on last phase of abstract representation it is possible to logging, that the information is accepted.
The information starts to prove and play a practical role in the nature only with the appearance of complex ES, transform entrance influences in codes, as reflections. Such ES possess the mechanism of own "supervision" from own side, which is a basis of ability of abstraction and consciousness occurrence. The effect of abstraction is formed by a real physical feedback mechanism, allowing system to "look" at itself from the side and to change a set of entrance signals owing to the memory.
Complex ES corresponding to level (Y(2)) are capable to carry out a generalisation and creation. If the ideal abstraction (2I) is included into a real life of an environment with the help of actuators, it exists in it owing to manifesting physical objects-carriers. It can be fixed on various carriers in the form of its properties.
Technology AmI is transparent, embedded in human surrounding and adapt to all his sense organs. AmI provides of simple interaction focused on the user, when high-quality information access and personalized content available to everybody, anywhere, and at any time.
The concept of AmI is strongly related with the long term vision of an intelligent service system whenever the technologies are able to embedding the platform in the required devices for powering context aware, personalised, adaptive and anticipatory services.
ES are "a computing kernel" of highly specialized systems, such as mobile phones, players, cash dispenses, various operating devises of cars and etc. Realization of the newest concepts of "clever houses", "smart dust", "intellectual highways", etc., completely will fill a life of the person.
Intellectual systems connected one with another set of wire and wireless connections; give good preconditions for closer integration into networks and creations of the full-function, simple and reliable connected to a network from any point. Such intellectual systems essentially differ from the existing equipment on a way of integration into the human environment and means of interaction with users.
Mentioned below results can be used as elements of methodology of construction complex ES and objects operated by them.
- By expansion functioning of the networks, consisting of set of elements, comes into existence macro level on which appears steady system characteristics. The macro level is not only a special way of the description of working system of elements, but it has the status of "existing reality". "Consciousness" is a kind of existence of system characteristics of the working elements organized in network.
- At the decision of a problem of construction intellectual systems, basic devices are sensor network, memory and ES with the set of functions, one of which is function of self-learning. In process of activity such system makes abstract reflection of the environment and creates its model. Thereupon, ES should solve the problems effectively.
The model exists owing to mechanism of high-organized network, and it cannot emerge separate for the ES and objects operated by it. Detection in such system, transformation of signals acceptance from a sensors network in model, will be the proof of beginning realization characteristics of consciousnesses in system.
- Dadunashvili S (2005) Coordinate system for representation of the Informational Phenomenon's. Georgian Engineering News, 2005, No.1, pp 36-47.
- Dadunashvili S (2008) Informational Phenomenon's in Embedded Systems, Conf. Proc., 21st International CODATA Conference, Kyiv, Ukraine, 2008, pp 218-225.