SBGames 2012

This weekend happened the SBGames 2012. The 11th edition of the largest Game Symposium in Latin America. The UnBiquitous was present with three of its members working on the organization. Professor Carla Castanho was the General Chair, professor Ricardo Jacobi as part of the local committee and Lucas Almeida in the staff.

The event was very far from other in the academy. The partial numbers included 1.600 participants and more than 15.000 visitors in the open area. The academy was pleased with many papers and games in three tracks (Computing, Art and Design, Culture and Industry), many keynotes from great names in the industry. For the general public was available an open area where game developers displayed their work to the community, art exhibitions and many other activities.

My part on this event was representing the first work from the UnBiquitous in the gaming area. My intent was to show how the expertise of professor Carla Castanho in Games has led our group into the area of ubigames. Our paper named "What's going on with ubigames" shows a review of published games on the last decade. From this study we extract a body of learned lessons and tendencies to and future developments. It was written in collaboration with Ana Ozaki, Ricardo Jacobi and Carla Castanho.

SBGames 2012 - What is going on with ubicomp games from Fabricio Buzeto

Apart of representing our study I was able to be present in all of the four tracks. This enhanced the view of games. Not only restricted to the papers presented, but also with the opportunity to share ideas with other researchers and developers.

UnBiquitous in a SBCUP 2012

In the latter half of 2011 I finished my graduate work called "TRUE - The system for tracking, locating and identifying users in smart spaces". This work was mentioned in another post on that blog.


As the previous post says this work was developed by me (Tales Porto) and Danilo Ávila. Currently I work at "Esec Segurança Digital" with mobile systems, web systems and cryptographic systems. Danilo Ávila work at "LG Sistemas" with a web systems.

Tales Porto

Danilo Ávila

This year, this work has yielded an article that was submitted to SBCUP 2012. The SBCUP 2012 or "IV Brazilian Symposium on Ubiquitous and Pervasive Computing" is a satellite event of the "XXXII Congress of the Brazilian Computer Society" which was held on July 18, 2012 in Curitiba - PR. 59 articles were submitted for the event, of which 16 were accepted. What is meant a rate of 27% of acceptance.

The presentation system can be seen below in a slidecast and article can be read at the following link (both in portuguese).

Apresentação do Sistema TRUE from Tales Porto

Augmented reality applied in the ubiquitous context

Excited about the work done by my colleagues in the group UnBiquitous, I thought of a proposal for the integration of augmented reality and ubiquitous environment to which the UnBiquitous was inserted. But first, let me introduce myself. My name is Ricardo Andrade, I am undergraduate student of Computer Science at Universidade de Brasilia and I currently work at CAIXA Information Technology with development in two areas: Data Quality and Corporate Risk, both involving DW (Data Warehouse), ETL (Extract, Transformation and Load) and BI (Business Intelligence). Additionally, I have worked for a long time with web development at Mirante Web Services.

Now, back to my work. Augmented reality offers a full complement of real world and virtual world, using elements or information of both worlds. This allows the visualization of new information, usually unseen to the naked eye, about real or virtual objects. This supplement brings real-time interactivity between both worlds, making the manipulation of information more natural.

The insertion of virtual objects may occur by two means: through markers or points of interest.

  

1. Markers: For the objects can be recognized as valid markers, they require a standard format consisting of a "frame", made with square edges in black. The identification code is located within these markers may be composed of symbols or a square array of black or white, as seen in the figures below.

             

After identification of the marker, starts the decoding process code inside. When this process is completed, a virtual object is associated with the identification code of the marker. Finally, the virtual object is superimposed on the marker.

In the example below, the marker was found from the frame and the car, the virtual object was associated with the code "Hiro" decoded and superimposed on the marker.

2. Point of interest: Unlike what occurs with the markers, points of interest does not have a pattern to be recognized. Identification and localization is based oncoordinates. The coordinates of location of the user is obtained using GPS, subsequently cross-positions of longitude, latitude and altitude mapped from a point of interest in a database.

Sights and places of interest can be mapped and for providing information according to user's profile, helping to find places of interest. When the camera is positioned to a point of interest, information can be presented as a description, distance, among others.

In the UnBiquitous, many works have been developed aiming at provisioning and using resources. With the aid of radar, we have information about what resources are available in the ubiquitous environment. However, the location of the devices on which the  resources resides, are not clear, making and intuitive view of the available ones a worthy problem.

My proposal consists of using augmented reality as a mechanism for viewing the available resources and providing a device redirection for a specific resource, based on the integration with the application Hydra. This application will be supported by the Android platform and is divided into:

1. Identification of markers: The identification code of the markers will be obtained by QRCodes, which represent the name of the  application running on the device. For the recognition of the marker and decoding QRCode party applications will be used ARToolkit and ZXing, respectively.

2. Finding the resources: The UOS middleware will provide a list of resources for the application.

3. Presentation resources: Here comes the augmented reality. Resources will be made available to the device and a virtual list overlapping the marker selected with the aid of the application DroidAR.

4. Redirection of resources: Finally, the user will be able to redirect a resource from the list presented to the Hydra application.

Advances will be given to the user achieve greater transparency and knowledge of resources available in the smart space and also ease the redirection of these resources.

Resource Classification for DSOA

Since 2010, we, Bruno and Marcelo, were determined to work with Ubiquitous Computing. Then we looked for professor Carla D. Castanho, who introduced us to Fabricio Buzeto, current project leader of Unbiquitous research group. After a few meetings, we agreed to develop a resource classification for DSOA architecture.

                  

Bruno Pessanha Marcelo Valença

Firstly let us introduce ourselves. We both are undergraduate students of Computer Science here at Universidade de Brasilia. I, Bruno Pessanha de Carvalho, currently work as a Systems Analyst at Oak Soluções, focused on software-based enterprise solutions and used to work as a Software Developer at Autotrac, focused on embedded software development. My partner Marcelo Valença de Almeida used to work as a Software Developer at Autotrac and Sea Tecnologia focused on mobile development and web development, respectively.

The problem that our research group was facing was that without the classification it was hard for the developer to find resources at the environment. For example, imagine that the user needs a nice and big screen to project his graduation presentation that it’s stored at his cellphone and there are three resources available: Wide40InchesScreen, Nice27InchesScreen and Small14InchesScreen. So, if the developer would like to find that three kinds of screen at the environment, he should search for that three unrelated identifiers. The development of a resource classification would make that user capable of querying just a Screen and then he would receive all the options that are equivalents to that type of resource. We could specify a Screen as a root of one equivalence tree, and then extends to WideScreen and LetterBoxScreen, as shown above.

For this, we had to study how the current standards (UPnP, IEEE 1451,DLNA, USB, Bluetooth) dealt with service subtyping. After that research, we were able to set our own standards: audio, video, image, keyboard, pointer and connectivity. On this basis, it became possible to define a hierarchy of extensible interfaces, allowing the selection of equivalent resources in order to ease the search for resources, thus increasing the adaptability of services, the focus of this project.

For one or more interfaces be considered equivalent, the following criteria must be served:

• the services must belong to the same resource;
• the application must have at least the same services as their "father" and they must have the same identifiers;
• they must have the same parameters.

Today the interface validation and the necessary modifications in DSOA are partially implemented and tested. Our focus now is on developing applications that serve as proof of concept to our implementation:

1 - Mouse composed of two buttons;

2 - Mouse composed of two buttons and scroll.

Thus, when some application request for resource (1), the resource (2) also maybe provided to meet the requesting application.

DSOA

The DSOA (Device Service Oriented Architecture) [1] consists of the guidelines on how the uOS operates in the smart space. Our purpose on defining the DSOA was to extend the characteristics of the SOA and to provide a architecture that lead to a way to organize a ubiqutitous environment. Our main goals are focused on three requisites (that are spread throughout all of our research):

1. Support limited devices.
2. Handle the interaction details between applications and resources.
3. Support platform heterogeneity.

This way the DSOA provides a model for the smart space. SOA suggests that all capabilities are available by applications and must be accessed through services. In an ubicomp environment the entities involved have more intrinsic relations between capabilities and how to provide them. In order to address this, DSOA presents an extension of SOA concepts for modeling a smart space in a more cohesive way. Besides, SOA does not address how interactions must occur, but ubicomp environments have some specific interaction details that must be handled. DSOA shows a set of strategies to deal with these situations.

Basic Concepts

Smart Space: “The smart space is a neither empty nor unitary set of devices provided with computing power and interconnected by a communication network in a collaborative way.”

Devices: “A device is a computing equipment with communication capabilities, which must host applications or make resources available in the smart space.” 

Resources: “A resource is a group of functionalities logically related. These functionalities must be accessible in the environment through pre-defined interfaces.”

Services: “A service is the implementation of a functionality made available in the smart space through a resource with a known public interface.”

Application: “An application consists of the implementation of a set of rules and behaviors related to the resources and users in the smart space.”

A representation of the concepts of a DSOA Smart Space

The above figure shows a representation fo such concepts but a example can better illustrate how they bind together:

• A smart space is defined as a living room. Its available devices are Dell PC (5), a Nexus One cell phone (6), a Sony LCD 52” TV (2) and a WebCam (3) connected to the PC, an air conditioner (1) and a motion sensor (4).
• The cell phone can turn available to the smart space its resources like camera, keyboard, microphone, speaker. Besides that the cellphone can host an application which allows the owner to interact with available resources.
• The camera resource can provide services like snapshot and video stream.

Roles: Following the definitions of the SOA, on the DSOA devices can perform according to three possible roles:

• A device act as a consumer when its accessing a resource provided by another device.
• Analogously, a device act as a provider when it is providing a resource to others.
• A device act as a register when it turns available for others its knowledge about the neighbor devices.

Communication Strategies: Focusing on how data is transported and how interaction occurs we group our strategies into two categories.

First we have the Data Transport strategies regarding on how data is exchanged between a consumer and a provider. On this group we find:

• Discrete messages are defined as the ones which carry information with known size. They are characterized by mutual knowledge of the beginning and the end of data in the communication. This strategy is used for small scale and discrete information, like requests for actions or queries.
• Continuum data are characterized by an information flow with no previous knowledge of its ending. The common usage of this type of communication is in data streams transmissions, like audio/video, or large scale data transfers.

Such communication can occur on any transport method (Bluetooth, Wifi, etc) but the DSOA defines logic channels to isolate each type of data.

• There is only one control channel where discrete messages are exchanged. This channel always exists between devices since its through it that service requests (control messages) are communicated. 
• Data channels are responsible for managing continuum data. There can be as many data channels as needed, since each of them enclosures a different kind of data stream that applications and resources must deal with.

On the other side, the Device Interaction strategies take care on how the resource access dynamics occurs. On this group we find:

• A synchronous interaction occurs when a consumer starts its service request and a provider responds in a request-response manner. Common uses of this type of interaction strategy can be found in command requests (like lowering the air-conditioner temperature) or in a simple information query (like asking for the current room temperature).

• Some interactions don’t start immediately after the client request. This type of interaction is defined as asynchronous and has the following steps. (i) A client asks a provider to be informed about the occurrence of an event. (ii) The provider receives the request and waits for the event to happen. (iii) In an arbitrary moment (which may never happen) the provider notifies the client of the occurrence of the event. This type of interaction is commonly used to identify changes on the environment like user actions or temperature changes.

Hope you liked our proposition. If you have any opinion leave us a comment and if you want to see a DSOA implementation just visit the uOS project.

Ending 2011

This post is just a Review of 2011 for our research group.

Summing up, it was a great year for us. Myself and Ana Ozaki were able to publish papers regarding our results in the uP and Ontologies. We hope this show more about our ways into the Ubiquitous Computing and were we want to go.

This year also two projects (par of the uOS) came to and end. One was the TRUE system, which added capabilities of User Tracking, Positioning and Recognition in the smart space. This was a great work that Tales and Danilo achieved. Other was the full implementation of the Hydra, a resource redirection application, conducted by Lucas Almeida.

For 2011 we have two new members to the group. Lucas Almeida and Carlos Botelho we'll be master students into our program. We also will have projects regarding improvements in the middleware itself and new applications.

So 2011 was great and 2012 will start full of great promisses.

WPOS 2011

Every year the Universidade de Brasilia (UnB) hold a workshop for all of its researchers. The name of the event is WPOS (Workshop de POS-graduação). Even though the focus is on Masters and PhD researchers, every year we also receive some very interesting results from undergrads.

This year was the first year that it was held in Brasilia itself, usually it's sited on a nearby city like Pirinópolis. Being so close is very fruitful for the ease of transport but don't come with the same atmosphere of engagement that the prior issues held.

During this edition the UnBiquitous was represented by myself (Fabricio Buzeto) and Ana Ozaki. Ana presented her results regarding Ontologies for smartspaces (in special de uOS). My job was to present my idea of a Code Mobility extension to the DSOA. The ideas were well accepted and received some valuable feedback.

Highlighting the some interesting parts of the day i can present: The talk from Professor Rohit Gheyi about its SafeRefactor tool. The always amazing speedup results from Alba Melo students regarding GPU and Clustering techniques and Deborah Mendes study regarding the W-Entropy Index for Social Networks.

And lets hope for the next year.