Conference starts in:

Use FRUCT discount code at

Find a new job

You are here


Background and Motivation: 

Smart space is an ecosystem of interacting computational objects on shared knowledge base. The key goal is seamless provision of users with information using the best available resources for all kinds of devices that the users can use in the ecosystem. This project develops demo smart space - SmartRoom, where the participants access services in the smart room using personal mobile computers. The room is equipped with sensor devices that sense the physical parameters of the environment and participant activity. All knowledge is collected, organized, shared, and searched in a common smart space. Local services run on local computers or nearby servers. The system accesses the external world for appropriate Internet services. Service outcome is visible online on the public screens or personalized on mobile clients.

Project Summary: 

The mission of Smart Room is to form a localized digital environment for holding conferences, meetings, lectures, trainings, and some other activities important for research, education, and business processes. Participants are located in a room and use the environment in their ongoing activity. The environment provides a set of services that simplifies activity organization and participation as well as automates technical functions of information acquisition, sharing, and transformation. The automated and intelligent support simplifies the human activity, allowing participants to concentrate on the problems the event is devoted to, not on technical details of information acquisition, sharing and transformation.

Smart-M3 based system architecture: Smart Room is a multi-agent intelligent system that provides a smart space for its agents. The smart room space (SR space) is implemented using the Smart-M3 platform. Agents communicate sharing their knowledge (data, semantics and any digitally encoded information) in the SR space. Information interoperability is achieved due to ontological representation models (RDF, OWL) and the smart space access protocol (SSAP, a part of Smart-M3). SSAP is XML-based and implements communication (on top of TCP) between an agent and the SR space (operations join/leave, insert/update/remove, subscribe/unsubscribe). SR space is maintained by Smart-M3 semantic information broker (SIB), which runs on a dedicated SR server. Agents run on smart room devices and operate as Smart-M3 knowledge processors (KP), i.e., they have own local knowledge, access shared knowledge in the SR space, apply reasoning over that knowledge, and make consequent decisions in accordance to the application logic (with possible publication of derivative knowledge in the SR space). Basic reasoning is performed on the space side and based on Semantic Web technologies (e.g., SPARQL queries attached to SSAP operations). 

IoT-aware environment: Smart Room is equipped by various devices (surrounding physical/digital objects), which are localized within the room or located nearby (e.g., in the server room of the organization). Each device can host one or more smart objects: autonomous everyday objects, being augmented with sensing, processing, and network capabilities, are transformed into smart objects that understand and react to their environment.

Service set: Agents (smart objects) running on SR devices are involved into service provision chains. Any service has a dedicated agent that performs functions on client demand and/or provides information related to specific component of the room. Some services (which can dynamically change their state between active and inactive) are available at the moment and are offered to participants in the room. Some services are public, i.e., intended for all participants and shown on large screens in the room. Services can also be accessed from a client agent installed on personal mobile devices. Although the services are available for all participants, a specific service subset is offered for each person based on his preferences and current context. It allows implementing personalization and enabling proactive delivery of the services.

Smart Room can operate in several modes, depending on the type of event holding in the room. 

1. Conference

2. Meeting

3. Lecture

Each mode has own core service set (partially overlapped with other modes). For instance, the conference set includes agenda and presentation. In any mode, Agenda describes the activity plan held in the room. For instance, a conference agenda is a list of talks (conference program). Presentation visualizes slides and other information of the current speaker. Agenda and Presentation have own screens in the room.

Each core service set can be augmented with advanced services.

1. World information services are accessed from the web. They provide information and processing facilities that can be used in the smart room. For instance, information trackers find appropriate knowledge collected in the web, e.g., presenter’s citation index from Google Scholar.

2. Discussion services allow online discussions between the participants. It uses public blog services available in the Internet. For instance, participants can discuss each other’s presentations during the conference or publish their opinion during the meeting.

3. Activity tracking services are deployed locally in the smart room. They accumulate knowledge appeared in the room and derive new knowledge. For instance, personalized reports sent to each participant after the conference.

4. Sensors are deployed in the room. They monitor parameters of internal physical environment and publish the measurements into the smart room space.

Project goals and future research directions: 
<p>The main research problems for the project are as follows:&nbsp;</p><p>• Development of design principles for automated ontology-driven programming for smart spaces (enhancement of SmartSlog SDK)</p><p>• Research for efficient Smart Room infrastructure on top of Smart Spaces and IoT: distributed architecture, ontological models, service-oriented approach (including personalization, adaptability and pro-activity), knowledge reasoning, multi-agent collaboration, hub-like application content repository.</p><p>• Problems and architectural solutions for cross-platform mobile client development: tradeoffs of cross-platform and platform-dependent development in mobile clients and development unification for various mobile platforms as much as possible.</p><p>• Definition of multi-service property and service personalization within mobile client UI.</p><p>• Multimodal UI for multi-user services: multiple information sources, reasoning over them, dynamic selection and composition of relevant knowledge in compact visual forms.</p><p>• Ontology-driven smart space content maintenance and knowledge reasoning and subject to space locality and dynamicity.</p><p>• Ontology-driven service construction and composition. Cross-service exchange. Self-construction of service chains for adapting, personalization, and proactive delivery.</p><p>• Efficient tracking of user’s activities and summarizing based on multiple local and external sources, including mobile data and ‘Big Data’.</p><p>• Mediation approach for synchronizing several Smart Rooms and for using Smart Room with other systems.</p>
List of team members and their organizations: 

Dmitry Korzun, Ivan Galov, Sergey Marchenkov, Andrei Vdovenko, Pavel Kovyrshin - Petrozavodsk State University

Project Timeline and Expected Deliverables: 

Jul.2012 - Dec. 2012:  Initial prototype based on Smart Conference System (from SPIIRAS)

Jan.2013 - May 2013: Experimental prototypes for services (sensors, blogging, mike, etc.)

Jun 2013 - Nov. 2013: Stabilization of the core service set.

Dec.2013 - Mar. 2014: Stable release and deployment at PetrSU.

Final deadline: 
Wednesday, December 31, 2014 (All day)