Goal Oriented Layered system for Inter-operable Activities of Things


CALL: 2013

DOMAIN: IS - Telecommunications and Multimedia

FIRST NAME: Benjamin





KEYWORDS: Classical networking aims in addressing individual hosts, by means of unique network identifiers, such as MAC or IP/DNS addresses, IMEI/SIM for GSM, etc. There is no global mechanism to link addresses with real meaning of a node; instead, one must know in advance the addresses of the nodes in order to use them. This is similar to the case of a phone call – one cannot make a call if he/she does not know the recipient’s number.Broadening the above example, while searching e.g. for an expert to fix a broken washing machine, one may use “yellow pages” – a semantic catalogue of companies and their phone numbers. This idea may be applied for networking – it would be desirable to comfortably access a function that is provided by a node from the network, with no knowledge about this node’s address. Moreover, in contrast to “yellow pages”, where we must manually choose one of the items (usually quite randomly, as we have no additional information how to compare the “competitive” entries), we would like to shift all the searching activities to the system, as this decision must be fast and as effective as possible. So, we would like to declare some global guidelines, to declare a ‘comfort” situation we want to place ourselves in. “Comfort” means something different for different people, situations, places, sometimes being a result of negotiations with other people, etc., and some “comfortable” conditions are sometimes defined in a fuzzy manner. Anyway, comfort covers global regulations, abstracting of capabilities of particular place/devices. Thus, defining and controlling such “comfort” is not a trivial task and should be left to dedicated software, possibly highly personalized and migrating across the network according to people (device) movements.Main goal of GOLIATH project is to propose an intelligent network capable of addressing functions (services) provided by the nodes (namely – Internet-of-Things IoT devices, computers/smartphones/tablets etc.) rather than the nodes themselves, based on the semantic descriptions of nodes’ capabilities and software agents for controlling the comfort in an individual manner. The proposal fits with a natural trend to treat intelligent systems as a single entity, rather than a set of individuals. Artificial intelligence is always related with networking – more intelligent is a system, more this system is treated as a single, consistent entity. This is usually a continuous evolution with final step – interacting with the “installation” rather than its particular elements such as a laptop, a mobile or a light switch.However, nowadays this is the individual access to the intelligent network nodes that is most common. In contrast, what we propose is to link the intelligence with the whole network, leaving to the system the tedious tasks related with proper device/node searching for, addressing, and accessing to gain certain functionality. “Intelligence” and “comfort” are to be related with the system, such as “intelligent home”, “intelligent office”, vehicle, hospital, etc.In a natural manner, the proposed system is divided into two layers. Lower layer, aimed in physical access (including addressing and information exchange) to IoT devices and other network nodes, makes it possible to represent nodes’ capabilities abstracting of their physical location, addressing and other identification means. This layer is implemented as an application of SITE framework, previously proposed for semantic management of IoT devices based on their functionality. The upper layer, realized as an intelligent MAS system, is responsible for defining, controlling and managing the comfort in a personalized manner. The above two-layer architecture makes it possible to address new application areas for ad-hoc and intelligent networking.

START: 2014-02-01

END: 2016-09-30

WEBSITE: https://www.list.lu/

Submitted Abstract

The aim of this project is to propose a network of IoT devices coordinately working for providing comfort at certain locations and for certain human users. As the meaning of comfort depends on individual preferences, and the capabilities of given location may vary, in general, comfort management is not a trivial task. Moreover, while several humans find themselves at the same location, the meaning of comfort should be negotiated and somehow averaged. The comfort is assured by activating functionality of IoT devices, first, for detecting current parameters (such as temperature, air flow, pressure, etc.), and then for switching on (or off) certain actuators (such as heaters, air-conditioning installations, etc.). The users are not aware of current capabilities of the set of IoT devices available at given moment and location. Instead, they declare guidelines for the behavior of the system, and these guidelines (goals) are mapped to devices’ functionality by the software agents. Software agents cooperate to assure comfort to the maximum extent. However, the obtained level of comfort may not be necessary equal to the one declared by the user. Such accordance depends on (1) other users in the neighborhood, and (2) possibilities of the set of IoT devices available.General architecture of the system aims in controlling comfort at two layers. First (upper) layer is based on software agents. This level is responsible for comfort management, including the storage and processing of user-defined requirements to personalize the meaning of comfort. Comfort definition is represented by software agents in the form of fuzzy-assessed goals. The meaning of comfort is mapped by the agents to the requests for activation of IoT devices. The requests are mapped to an activation of a set of these IoT devices that are the most suitable to fulfill these requests. Main project objectives are the following:•providing efficient way of modeling of the comfort and its parameters, assisted with the negotiations to equalize comfort meaning among multiple users,•defining mapping procedure from comfort descriptions and fuzzy requirements to generic requests of activation of IoT devices,•designing efficient way of dynamic activation of given IoT devices based on incoming requests and semantic descriptions of devices’ capabilities.The proposed systems assures:•fulfilling the goals declared by the users to the maximum extent, usually restricted only by the capabilities of local set of IoT devices and contradictory requirements of “neighborhood” users;•portability of goals, requests and devices’ capabilities;•dynamic configuration of a local set of IoT devices, available at current location and moment of time;•a possibility of ad-hoc interaction with the system at unknown in advance locations and in unpredictable situations.Perspective applications of the proposed system are the following: improving comfort of life by mass usage IoT devices, also in the ad-hoc manner, improving energy usage by coordinating energy consumption within households and workplaces, public places, institutions, local administration, etc. However, to restrict the scope of this proposal, we concentrate on the application area related with “intelligent” homes and workplaces, i.e., inside a building. We plan to verify the idea as a real-case scenario by means of Neobuild company and their system for “intelligent home”.

This site uses cookies. By continuing to use this site, you agree to the use of cookies for analytics purposes. Find out more in our Privacy Statement