Communicating Models for Smart Cities Applications
The concept of Smart Cities has become an important asset being considered by several countries to make their cities more intelligent to improve the quality of life of their populations. Applications within the context of Smart Cities range from Healthcare, Urban Mobility, Water Resources capture and distribution, Education, Agriculture, Environment, Emergency deployment of access to internet, etc. Such applications depend on both hardware and software, as well as sensors and IoT. The proposal described here discusses the use of a specification model to represent each and every Smart Cities system covering the applications. Statecharts can represent reactive systems with the concepts of hierarchy, parallel activities and synchronization. So, they can be employed to represent Smart Cities Systems independently. However, as there is a lot of interaction among such applications, it is relevant to devise mechanisms of communication among independent applications. For example, representation of such systems enable validation, verification or performance evaluation. In order to develop these communicating models, it is necessary to conduct a thorough study to evaluate if there is any necessity of extending the syntax and semantics of Statecharts or to describe a novel approach for said integration. Moreover, one has to consider that systems could be continuous or discrete and how to deal with their interactions. Therefore, some studies and implementation of a case study of such communication are in order to develop this proposal.
Natural and Technological Disaster Management through Geospatial Artificial Intelligence (GeoAI)
This proposal is towards Ph.D. Degree in Applied Computing (CAP) at INPE. It is part of the project IARA Recreating Urban Spaces approved by the Call FAPESP-MCTI-CGI.BR
This line of research is part of the IARA project – Artificial Intelligence in the Remaking of Urban Spaces, which has, among its objectives, the application of Artificial Intelligence techniques to solve issues related to the smart cities theme and improve the quality of life in urban environments.
One of the crucial aspects when considering smart cities is the need to have access to organized and good quality geospatial information that contributes to different application domains. On the other hand, we currently have a huge amount of daily data, particularly matrix data, and we often need to respond in real time to some critical situations that occur in urban environments. In other words, real-time data and AI-based interpretations are two of the pillars for meeting the needs of the cities of the future.
In recent years, a sub-area within Artificial Intelligence called GeoAI (Geospatial Artificial Intelligence) has emerged that seeks to combine methods related to geotechnologies (such as Geographic Information Systems and Remote Sensing) with artificial intelligence and data mining). In this way, GeoAI can be characterized as a holistic geospatial view integrating vector, matrix, drone, sensor (IoT) and drillhole profile data in the same environment allowing for analysis, monitoring, modeling, and data management to produce short-term scenarios or long term or resolving issues in real-time.
Our aim within the IARA project is therefore to explore GeoAI applications to generate new insights from geospatial information by developing or applying algorithms to detect and classify features with a much greater degree of accuracy and speed than was possible with the traditional methods. Among the research development areas, we can highlight two: (a) developing machine learning methods and tools to address geographical problems and discover new knowledge from geo(big)data; (b) apply deep learning to large 2D and 3D geospatial databases.
When applied to cities, GeoAI enables geospatial analysis and modeling that can contribute to a variety of critical activities in an urban environment. Our main emphasis is related to issues related to the management of natural and technological disasters that can impact urban areas involving disciplines such as geology, geomorphology, pedology, hydrology/hydrogeology, climatology, geotechnics, and land use. According to different IPCC scenarios, the situation of natural disasters tends to accelerate due to greater climate variability with an increase in maximum or minimum precipitation events with serious implications for densely occupied areas. Among the applications that will be emphasized include flood mapping, monitoring and analysis of mass movements, management, and monitoring of risks in dams, collapse and subsidence by natural and anthropogenic effects, and characterization of escape routes associated with different types of disasters.
Code and Test generation from UML Models
The development of real-time embedded software demands, such as Space Applications and Smart Cities Applications and other critical applications, compliance of specific requirements. During the development of such software, especially in the initial stages, performing verification can come in handy to evaluate whether the requirements will be met. This aspect can significantly increase the quality of the system and its control. Considering the increasing use of model-driven development, system analysis techniques are necessary.
The application of MDE (Model-Driven Engineering) techniques requires reasoning in the design of the systems. The quality of the generated system is directly related to the quality of the produced model. In order to improve the quality of these models, we seek to apply validation and verification techniques to the models. With an embedded system specified in Statecharts using UML diagrams, very significant information about the future behavior of the system can be extracted. In this M.Sc. proposal, activities will be focused on automatically generating not only test cases but also target system code from UML diagrams. These test cases can then be used to test whether the designed system behaves as expected. It is expected to apply and develop techniques for model to model transformations and models to text transformations.
The site was designed with Mobirise