Integrating Cyber-Physical Systems in a Component-Based Approach for Smart Homes.

Integration of different cyber-physical systems involves a development process that takes into account some solutions for intercommunicating and interoperating heterogeneous devices. Each device can be managed as a thing within the Internet-of-Things concept by using web technologies. In addition, a “thing” can be managed as an encapsulated component by applying component-based software engineering principles. Based on this context, we propose a solution for integrating heterogeneous systems using a specific component-based technology. Specifically, we focus on enabling the connection of different types of subsystems present in smart home solutions. This technology enables interoperability by applying a homogeneous component representation that provides communication features through web sockets, and by implementing gateways in proprietary network connections. Furthermore, our solution eases the extension of these systems by means of abstract representations of the architectures and devices that form part of them. The approach is validated through an example scenario with different subsystems of a smart home solution.

OntoTrader: an ontological Web trading agent approach for environmental information retrieval.

Modern Web-based Information Systems (WIS) are becoming increasingly necessary to provide support for users who are in different places with different types of information, by facilitating their access to the information, decision making, workgroups, and so forth. Design of these systems requires the use of standardized methods and techniques that enable a common vocabulary to be defined to represent the underlying knowledge. Thus, mediation elements such as traders enrich the interoperability of web components in open distributed systems. These traders must operate with other third-party traders and/or agents in the system, which must also use a common vocabulary for communication between them. This paper presents the OntoTrader architecture, an Ontological Web Trading agent based on the OMG ODP trading standard. It also presents the ontology needed by some system agents to communicate with the trading agent and the behavioral framework for the SOLERES OntoTrader agent, an Environmental Management Information System (EMIS). This framework implements a "Query-Searching/Recovering-Response" information retrieval model using a trading service, SPARQL notation, and the JADE platform. The paper also presents reflection, delegation and, federation mediation models and describes formalization, an experimental testing environment in three scenarios, and a tool which allows our proposal to be evaluated and validated.

A virtual-based task to assess place avoidance in humans.

Spatial memory can be assessed by virtual reality-based tasks, some of them similar to those previously used in rodents. In this study a virtual place avoidance task was developed based on a rodent labyrinth to evaluate human spatial memory. The task required subjects to avoid in a virtual room a prohibited zone defined solely by distal cues, while they collected rewards. The size of the to-be-avoided area was used as an independent variable as well as the gender of the participants. Subjects displayed an adequate performance after a single 5-min trial, entering the prohibited area on very few occasions. Our results showed that the size of the prohibited zone can be used to adapt task difficulty to different populations. Moreover, males were faster than females. They collected more rewards and spent less time in the prohibited area. Data collected supports the spatial component of the task and the validity of this procedure for spatial memory assessment in human beings.

A new virtual task to evaluate human place learning.

This study assesses the effectiveness of a new virtual task to evaluate human place learning. This test was based on the hole-board maze, developed for rodent research. Its design provides an easy set of levels of difficulty. Sixty-three undergraduate students (30 men and 33 women) were randomly distributed into three testing conditions; they had to find 3, 5 and 7 rewards, respectively, in a virtual room with 16 possible rewarded positions. Subjects were asked to use the minimum amount of attempts to discover all the rewards in 10 trials. In the initial trial subjects needed to visit almost all the positions to discover the rewards. However, in the last trial an important percentage of subjects did not err. Results showed that all subjects acquired the task but with different amounts of mistakes directly related to the level of difficulty of the condition. In addition, women were slower and less accurate than men. These results agree with previous results in other virtual tasks, and support the spatial component of this test.