This project operates in applied research in different areas of Computer Science, with a bias in the development of computational artifacts. In particular, the research undertaken in this project appropriates techniques, approaches, and methods of Data Science for the development of such artifacts. The treatment of the large volumes of data generated by the artifacts produced in this project are inputs for the analysis, interpretation, or even prediction of behaviors. Although it has a generalist character, most of the research is focused on teaching support, the construction of intelligent multimedia environments and decision making on the Internet of Things (IoT).
In the context of teaching support, we seek to build a cognitive and playful set of knowledge, through the development of artifacts that allow an increase in productivity in the teaching-learning relationship. These artifacts aim not only to verify learning but also to provide an environment that values students’ experience and experience, working with aspects of their worldview. The main multidisciplinary instrument for this development is robotics, which brings together knowledge of electronics, automation, computer science, mathematics, semiotics, and cognitive knowledge. Thus, tools are developed based on artificial intelligence, robotics, with consequent qualitative analysis of measurement data of these applications, with the students, in order to measure the adherence of tools developed along the educational and social aspects.
In intelligent multimedia environments, it is sought to create systems, whether embedded or not, that allow the execution of innovative applications focusing on the synchronization between traditional media objects, sensorial stimuli and multimodal interaction. Data science permeates the construction of these systems, providing a basis for interpretation of sensor data, detection of multimodal interaction, the perception of user state, as well as the use of traditional media objects with a differentiated approach. Also, traditional multimedia concepts are used to create models and authoring languages with the purpose of specifying intelligent environments and facilitating the specification of so-called multisensory applications (Multisensory Media), as well as the validation of its operation. Abstractions are proposed aiming at the representation of the interpretations obtained with Data Science; the sharing of information between sensor devices and actuators; and the independence of such applications from the execution platform.
Finally, in the context of IoT, the goal is to develop systems that can monitor the flows and behavior of individuals and vehicles, as well as identifying them automatically or that can perform the monitoring of climatic conditions, providing measures that can prevent environmental tragedies. Such systems must be efficient both in data traffic and energy autonomy requirements, besides being evaluated in the safety, validity and correction aspects. The projects developed by this area include the concepts of Intelligent Cities and Industry 4.0, among others.
- Diego Brandão
- Eduardo Ogasawara
- Gustavo Guedes
- João Quadros
- Joel dos Santos (Leader)