CC BY
89
TCYBER-PHYSICAL OBJECTS AND THEIR MODELS FOR MULTISENSORY DATA REPRESENTATION
Gleb G. Rogozinsky,
The Bonch-Bruevich St.Petersburg State University of Telecommunications, St.Petersburg, Russia; Solomenko Institute of Transport Problems of the Russian Academy of Sciences, St.Petersburg, Russia, gleb.rogozinsky@gmail.com
Keywords: cyber-physical systems (CPS), cyber-physical space, multi-domain model, multisensory, sonification, modeling.
The multisensory representation of cyber-physical systems (CPS) in human-machine interfaces of monitoring demands developing a set of new classes of CPS models. The desired CPS model classes should be focused first on reflecting the duality of cyber-physical objects (CPO), i.e. flexible representation of both physical and cybernetic (informational) components of the objects under monitoring. Also, the new CPS model classes should represent the features of cyber-physical space in a way they can be well adapted to the upper levels of monitoring infrastructure, i.e. the sonification layer at the edge between informational domain and cognitive domain.
In this paper we propose four classes of cyber-physical models, each of which takes in account some important characteristics of an analyzed system. Together a set of proposed models can be used to describe any complex system in common terms, freeing the upper levels of the monitoring infrastructure from redundant details and informational noise.
Information about authors:
Gleb G. Rogozinsky, PhD, Deputy Head of Medialabs, The Bonch-Bruevich St.Petersburg State University of Telecommunications, Senior Researcher, Solomenko Institute of Transport Problems of the Russian Academy of Sciences, St.Petersburg, Russia
Для цитирования:
Рогозинский Г.Г. Киберфизические объекты и их модели для мультисенсорного представления данных // T-Comm: Телекоммуникации и транспорт. 2017. Том 11. №11. С. 68-73.
For citation:
Rogozinsky G.G. (2017). Tcyber-physical objects and their models for multisensory data representation. T-Comm, vol. 1 1, no.1 1, рр. 68-73.
T-Comm ^м 11. #11-2017
7T>
1. Introduction
Recently we can witness the outbreak of the Cyber-Physical Systems (or CPS), as a main carcass of an upcoming Industry 4.0. The CPS concept tends to unify the domain of physical objects and their so-called digital shadows or virtual representations, which exist in the informational (cybernetic) domain. Joined together, the two domains form a Cyber-Physical Space. Above it lays the cognitive domain, where the human beings are able to extract knowledge from the information they receive through their bio-sensory systems (typically audiovisual ones), and make a decisions to control the systems. Taking it in account, we separate the interaction between hardware objects, software entities and humans into three domains, which are Physical Domain (PD), Informational Domain (ID), and Cognitive Domain (CD). Such approach is known as Multi-Domain Model (MM), first described by SotnikoV [1], [2]. The author used the MM to create a Modified Multi-Domain Model (or 3M), in order to get an abstract description of data representation in mono- and multi-sensory systems [3J.
2. The Main Features of Mnlti Domain Model of Communications
According to the MM and 3M concepts, any kind of activity can be understood as an exchange between three main domains -PD, ID and CD.
The PD is typically concerned with the energy processes and the interaction of material world objects. The situation analysis and intellectual activity are the products of mental and psychic activity of the CD. The ID is the area for the circulation of data used in the CD, representing the objects, phenomena and process of the PD.
At the domain borders the corresponding interfaces perform the information interaction between different elements of the system. Each object/subject of the system can be characterized by the finite number of states, represented by it own thesaurus.
Thus, the object A of a PD with corresponding thesaurus states, mapped onto the multiplicity of informational representa-
(4
tions
* of the thesaurus C, A
(1)
In other words the information is transferred, when the signal transmitting an image (notion) from the varifold thesaurus of the source system A into the varifold thesaurus of the target system B is changed.
The information is received when a new image of the source is formed within the varifold thesaurus of the target system.
.....s*
a
(2)
where Q^Q2 ~ mapping operators between different domains, i.e. PD, ID or CD; thesauri of a target signal and a source
signal.
In detail, we can write (2) as following
Wf
(3)
Thus the mapping between the corresponding domains is actually the operation of information impact between entities of the domains, expressed in the discovery of maximum conformity between elements of thesauri ¿j ,
Information impact is the influence of the "source" A on the state of the "target" B, which manifests itself in the change of the image { B}, seen in the variety of elements within the thesaurus
of the "source" ( B)%A. Since the "source" £,A and "target" 4b thesauri are different, the original image in the internal thesaurus
of ( B)%B and the image of { in the "source" thesaurus are also different. This can lead to errors, or inaccurate representation of the object in the thesaurus of the user.
Information exchange is the receiving and transmission of
signals leading to the mutual alteration of images {A and
{ of the exchange participants. This can be caused by alteration (expansion) ofthe participants' thesauri 4a and 4b-
Information interaction is the mutual change of images of
own systems of (A } - 1 and ( leading to the change of images ( A W and { in the other participants.
Information system (IS) is a system containing "information" and providing it to the user, A necessary condition is as follows: "The necessary components of an IS are: the user and the potential information". A sufficient condition is "The user and the potential information form an IS". IS are made up of elements which are information images (A) $f the real (material and immaterial) entities^ and possess information significance.
Information significance is the property of representing the entity, which requires a descriptive method containing a set of basic "meanings", immanent to the entity. "Information significance" should not be confused with "value" or "usefulness"; these properties are related to the users and the possibility of satisfying their need of information.
The formalized set of"meanings" is the 4a thesaurus. The entity item has a number of discernible states, which are perceived by the observer as a set of images of an object, each having its own "meaning". The number of slates determines the potential information carried by the object. When the observer acquires an image ofthe object (by means of perception and recognition), the potential information is actualized on the basis ofthe information representation ofthe object. The potential presence of information in system A is determined by the set of discernible states of the system and the varifold system thesaurus q,,.
Perception ofthe information transferred occurs when the receiver R acquires a new image of system A in the varifold receiver thesaurus {{A
User U is a person, object or process capable of perceiving
images (S)^ and possessing its ow n receiver thesaurus i,f; [3].
3. Issues and Limitations
of Traditional Data Representation
The problem of monitoring remains active even in the era of BigData. To add more, the ability to have an access to logging of literally anything at any time and anywhere, i.e. the output of any process and status, determines new demands and problems for the up-to-date monitoring and controlling systems. The increasing informational volume issues inspire new developments in a design field of human machine interfaces, since the most common and straight-forward solutions, typically based on visual representation of any received data, turn to be ineffective in cases of simultaneous visualization of thousands of objects. So to say, under conditions of the informational overload we should research alternative ways to represent and understand the constantly rising volume of information.
The human race typically exploits graphical interfaces to represent any data, although other solutions are possible. It includes various sonification methods (representation of data with nonspeech sound) [4J, tactile sensing and olfactory interfaces. The auditory representation of data has a long history [5], including altimeters and Geiger counters, although most systems of such kind exploit sound sets of rather small number of symbols. The development of sonification system for the modem informational world systems demands first creation of methodology, based on system analysis, sound design, acoustics and data mining.
It is often difficult for operator of some complex system to perform an adequately full and quick analysis of raw big raw through using only a set of visual toolkits. Traditional methods of modeling are not always suitable for obtaining necessary knowledge contained in the arrays of big data.
Any attempts for visualization of big data, such as representing of various complex networks and the included data as graphs, are useful in some way, but reveal the main drawback of that class of methods - the use of the v isual sensory system only. The human eye is able to distinguish about ten million colors, but it can not keep in focus many parameters at the same lime.
The architecture of a visual presentation of a complex system is not always able to correctly choose the optimal field of view -or requires special additional development (revision, additional fixing, tuning). Software model tools can excessively modify initial array when rendering it.
The specificity of monitoring processes of complex systems creates an issue of development of fundamentally new interfaces for displaying big data. Regarding the auditory interface along with the visual one seems to be self-evident but is also provable by practice and science [6].
4. An Approach to Cyber-Physical Space Representation
Cyber-physical systems as a generic concept for a vast variety of different complexes and infrastructures demand some generic approach to transfer the data into cognitive domain through multi-sensory representation. The developers of specialized software and hardware solutions for monitoring demand the unified approaches, as generic as possible to shorten the development time of the monitoring and controlling solutions, as well as related expenses and the level of uniqueness. Also, the solutions based on unified approaches will provide flexibility in adapting to infrastructinal changes, scaling or encapsulation of existing
solutions into bigger Systems of the upcoming future of technological synergy.
Thus, the specialized class of models to be proposed, providing connection between the cyber-physical objects (CPO) and symbolic space at the edge between informational and cognitive domains.
We use the concept of a cyber-physical space to define a common space, which include both physical objects and cybernetic (informational or virtual) entities, also referred as digital shadows or digital ghosts. Since one of the main logic line of Industry 4.0 proclaims congregation of physical objects and their virtual models, the cyber-physical space concept of physical and informational duality allows to represent those two parallel universes as a single world.
As it was staled in the beginning we focused mainly on mul-tisensory representation of data. So besides of some existing ey-ber-physica! space and systems models, we should propose our own set of models.
Below we propose several classes of models, eveiy of which describes some feature of a CPS: a separate CPO, the topology of a network of cyber-physical objects, a cluster of CPOs, and a hierarchical abstraction. Joined together the entities of those classes define any CPS in common way, which lets to develop mono- and multisensory interfaces of data representation at high level of abstraction.
5. Four Cyber-Physical Models
Here we give four CPS models, starting from the most simple, which we call a Cyber-Physical Atom.
Cyber-Physical Atom
The first cyber-physical model we introduce is a Cyber-Physical Atom (CPA-model, or simply A-model). It should become the minimal element of a cyber-physical space, and represent the duality of CPS. It has the physical part (Ph) and the cybernetic pan (Cy). Both of them together form the common entity, which possesses to both physical and informational domains. Such approach allows understanding the cyber-physical object as a composite, but nevertheless a single entity.
An interesting parallel can be found in mathematical complex numbers, where each number may be presented by its real and imaginary part. The given analogy to complex numbers assumes that we may have physical-only object, or visa versa, virtual-only object. The first can be incorporated into CPS by interacting with some other object, which has its digital shadow, thus the result of interaction is translated into cyberspace. The virtual-only entities are just software products, which can also interact with other software and some of them can be able translate its state to physical world though various actors. The reversal process of turning physical states into informational ones is carried thought using different kinds of sensors.
Assume that X is a Cyber-Physical Atom. When (x"f is CPA described in its own complex thesaurus^. We introduce two operators - Ph and Cy - to define the physical part of X (object X in a physical domain) and the cybernetic one as:
Ph
i
Cy
is a physical object, or a part of a CPA defined by its PD thesaurus, w hile (x)^*) is a cybernetic entity,
or a part of a CPA defined by its ID thesaurus.
The Figure 1 gives a schematic representation of a CPA model. Acting is an action of a direct impact from ID towards PD, while Sensing is a corresponding action form PD to ID.
PD -1|- ID
Fig. I. The Cyber-Physical Model I: Atomic Model
We need the described A-niodel to define and represent the behav ior or state of some stand-alone CPO, for instance in case we would like to separate or outline some exact CPO in the cy-ber-physical space. Meanwhile, typical CPS is a network, in which the smart things are interconnected, so the single CPA should be taken further as a separate node of a network.
(or more) elements of a physical world, e.g. the robot's hand hold the spray can and covers an aircraft element with the paint.
To compare with previous two operators, the macro-operator QP_t is an inter-domain operator, i.e. it describes the cyber-physieal impact. We should notice lhat such impact is a one-way only, i.e. it exists when some CPA 'senses' the physical part of another CPA. Further, we can separate inner sensing (or sensing which takes place inside the single CPA), and externa! sensing (when one CPA, which is actually its cyber-part, 'senses' the physical part of another CPA). The Figure 2 gives a graphical representation of the model.
Cyber- Physical Topologj>
Our second model is a Cyber-Physical Topology (CPT-model, or simply T-model). We include it to account the network features of a CPS.
The CPT-model defines the connections and different communication cases between separate CPAs. The main feature or representing descriptor of T-model is a connectivity matrix, or a corresponding graph model. The main diagonal of the matrix should contain Is. Connections between i and j CPAs of reviewed topology is defined by operators Q*' , Q "' H Q "c. Zero elements inside the matrix mean the absence of connections between corresponding CPAs.
The Q operator describes cybernetic-only connection, or the thesauri exchange, which takes part completely in the ID. Such operators can be also named as macro-operators, since they can implicit multiple transformations of different kinds,
Q ~ Q* Q1" ■■' Qn •
The cybernetic or informational exchange is performed by different network protocols. The discussed models do not describe the actual exchange conditions, but rather provide the way to describe the entities and structures of entities existing.
Respectively, the macro-operator QP1> is a physical-only connection operator, i.e. it describes the thesauri exchange taken place only in PD. We mean the direct interaction between two
The T-models together with a corresponding set of A-models are able to fully describe any system, although we may also need to incorporate couple more abstractions to simplify the process of mapping and the representation thesaurus design.
Thus the next model we discuss is a clustering model. We need it because of the high scalability of recent systems. No one can estimate the size limits for some existing system. The CPS can be integrated into bigger CPS; the later can scale up possibly to some cyber-physical mega-structures. Also it can be useful to review some constellation of CPAs as a single entity. In this case it becomes a question of a scale.
Cyber- Physical Cluste r
The third model in the CPO is the Cyber-Physical Cluster (C PC-mo del, or C-model). As it was mentioned above, C-model is an abstraction above the constellation of several CPAs. It can resemble the topology of included networks, but it is main feature is a common thesaurus similar to the thesaurus of a single Amodel. So the C-model thesaurus is a sum of all CPAs thesauri (Figure 3).
We may want to review the whole area of smart factory as a single cyber-physical element.
T-Comm Vol.11. #11-2017
m
КИБЕРФИЗИЧЕСКИЕ ОБЪЕКТЫ И ИХ МОДЕЛИ ДЛЯ МУЛЬТИСЕНСОРНОГО ПРЕДСТАВЛЕНИЯ ДАННЫХ
Рогозинский Глеб Гендрихович, Санкт-Петербургский университет телекоммуникаций им. проф. М.А. Бонч-Бруевича; Институт проблем транспорта им. Н.С. Соломенко Российской Академии Наук, Санкт-Петербург, Россия,
gleb.rogozinsky@gmail.com
Дннотация
Полисенсорное представление киберфизических систем в человеко-машинных интерфейсах обуславливает проектирование новых классов моделей киберфизических систем. Искомые классы моделей должны отражать дуализм, присущий киберфизиче-ским объектам, т.е. обладать способностью обеспечивать гибкое представление как для физических, так и кибернетических информационных) компонент наблюдаемых объектов. Новые классы киберфизических моделей должны отображать элементы киберфизического пространства для верхних уровней инфраструктуры мониторинга, например, для слоя сонификации на границе информационного и когнитивного доменов.
Предложены четыре класса киберфизических моделей, каждый из которых учитывает соответствующие значимые характеристики рассматриваемой системы. Совместно предложенный комплекс моделей позволяет описывать различные системы в общих терминах, освобождая верхние слои инфраструктуры мониторинга от избыточных данных и информационного шума.
Ключевые слова: киберфизические системы (CPS), киберфизическое пространство, мультидоменная модель, полисенсорность, сонификация, моделирование.
Литература
1. Сотников А.Д. Принципы анализа прикладной области в инфокоммуникационных системах здравоохранения // Тр. уч. заведений связи. 2004. № 171. С. 174-183.
2. Сотников А.Д. Классификация и модели прикладных инфокоммуникационных систем // Тр. учеб. заведений связи. 2003. № 169. С. 149-162.
3. Сотников А.Д., Рогозинский Г.Г. Мультидоменная модель инфокоммуникаций как основа построения аудиальных интерфейсов для мультимедийных информационных систем // T-Comm: Телекоммуникации и транспорт. 2017. № 5. С. 77-82. (на англ).
4. Германн, Хант, Ньюхоф. Книга о сонификации. Logos Verlag Берлин, 2011. (на англ.)
5. Уоралл, Д. Введение в сонификацию данных / Оксфордские труды по компьютерной музыке. Под ред. Р. Дина, 2009. 624 с.
(на англ.)
6. Рогозинский Г.Г., Лыжинкин К.В., Егорова А.Н., Осипенко И.Н. Программный комплекс сонификации динамических графов // Труды ЦНИИС. Санкт-Петербургский филиал. 2(2). 2016. С. 26-32.
Информация об авторе:
Глеб Гендрихович Рогозинский, к.т.н., зам. начальника НОЦ "Медиацентр", Санкт-Петербургский университет телекоммуникаций им. проф. М.А. Бонч-Бруевича, с.н.с., Институт проблем транспорта им. Н.С. Соломенко Российской Академии Наук, Санкт-Петербург, Россия
