Víctor Zamudio

Understanding and avoiding interaction‐based instability in pervasive computing environments

Purpose – This paper aims to address a fundamental problem related to the interaction of rule‐based autonomous agents in pervasive and intelligent environments. Some rules of behaviour can lead a multi‐agent system to display unwanted periodic behaviour, such as networked appliances cycling on and off.Design/methodology/approach – The paper presents a framework called interaction networks (INs) as a tool to describe and analyse this phenomena. In support of this, and as an aid to the visualisation and understanding of the temporal evolution of agent states, a graphical multi‐dimensional model (MDM) is offered. An instability prevention system (INPRES) based in identifying and locking network nodes is described.Findings – Both IN, MDM and INPRES enable system designers to identify and prevent cyclic instability. The effectiveness of the approach is evaluated using both simulated and physical implementations.Research limitations/implications – The problem of cyclic instability is strongly related to the num...

A Comparison between Metaheuristics as Strategies for Minimizing Cyclic Instability in Ambient Intelligence

In this paper we present a comparison between six novel approaches to the fundamental problem of cyclic instability in Ambient Intelligence. These approaches are based on different optimization algorithms, Particle Swarm Optimization (PSO), Bee Swarm Optimization (BSO), micro Particle Swarm Optimization (μ-PSO), Artificial Immune System (AIS), Genetic Algorithm (GA) and Mutual Information Maximization for Input Clustering (MIMIC). In order to be able to use these algorithms, we introduced the concept of Average Cumulative Oscillation (ACO), which enabled us to measure the average behavior of the system. This approach has the advantage that it does not need to analyze the topological properties of the system, in particular the loops, which can be computationally expensive. In order to test these algorithms we used the well-known discrete system called the Game of Life for 9, 25, 49 and 289 agents. It was found that PSO and μ-PSO have the best performance in terms of the number of agents locked. These results were confirmed using the Wilcoxon Signed Rank Test. This novel and successful approach is very promising and can be used to remove instabilities in real scenarios with a large number of agents (including nomadic agents) and complex interactions and dependencies among them.

A multi-dimensional model for task representation and allocation in intelligent environments

In the future, homes will have numerous intelligent communicating devices, and the user would like to configure and coordinate their actions. Appliances and people in intelligent environments will have some degree of mobility. If the user wants to go from one place to another, using the same community, the agent should be able to generalize the service, trying to build an equivalent collection of coordinating services. This ‘work in progress’ paper addresses this issue by proposing a multi-dimensional model that allows visualistation of devices, temporal relationships, mutual interdependencies and the environment dynamics. The model both offers a simplified means of visualising the task space and the interdependencies together with a means of reasoning about algorithmic solutions to task processing. The work is aimed at supporting research into Pervasive Home Environment Networks (PHEN) which is funded by the UK’s Department of Trade and Industry Next Wave Technologies and Markets programme.

c-INPRES: Coupling Analysis towards Locking Optimization in Ambient Intelligence

Ambient Intelligence, and in general, any autonomous rule based system has been found to suffer from cyclic instability. This behaviour is characterized by unwanted oscillations, due to interacting rules within networks of pervasive computing devices. The binary behaviour of each agent is defined via a set of boolean rules, and the behaviour of the system as a whole is given by the ensemble of rules defined over the set of agents. From complex theory it has been found that the problem of cyclic instability cannot be solved analytically; however, it is possible to prevent it. In this paper we present a novel solution based on locking, to prevent cyclic instability. This strategy makes use of the topological properties of the digraph associated called Interaction Network (IN), and the local rules of the interacting agents. The concept of strong and weak coupling is introduced. Using the strong and weak concepts, a strategy c-INPRES that minimizes the number of agents locked is presented. Preliminary and encouraging results are shown.

Physical Implementation of a Customisable System to Assist a User with Mobility Problems

The implementation of technology at home has caught the attention of many people, especially if elderly, handicapped or disable people are living on it. It is important to consider that the technology must be friendly with users, and even adapt to their needs and desires [1]. In this research, we present the physical implementation of a general system to assist users and patients in daily activities or duties. The system was implemented in a room and include sensors, actuators and wireless agents. This research is focused on people with some movement restrictions. The results are very promising, the user was successfully able to move several elements in the room (like windows or blinds) using only a mobile device as a tablet or a smart phone. Additionally, in order to see his potential adaptation and versatility, several devices where connected successfully to give comfort to the users.

Design of a Middleware and Optimization Algorithms for Light Comfort in an Intelligent Environment

The evolution of technology allows to people with special capabilities of mobility to perform the activities faster and easier. The intelligent environments combined with optimization algorithms and middleware agents could help to this aim. This paper presents the design and the implementation of an architecture of a middleware agent that allows us to make the communication between heterogeneous devices (sensors and actuators of different communication protocols from WiFi to ZigBee). On the other hand, we present a comparison study between micro-algorithms used to get lighting comfort in order to perform an activity in a confined space; this is affect by the light from the outside, which can be blocked by shutters and doors, and lighting of lamps obtained within this space. The micro-algorithm evaluated were: Genetic Algorithm (GA), Artificial Immune System (AIS), Estimation Distribution Algorithm (EDA), Particle Swarm Optimization (PSO), Bee Algorithm (BA) and Bee Swarm Optimization (BSO).

Intelligent System for Learning of Comfort Preferences to Help People with Mobility Limitations

It is essential to provide better living conditions for vulnerable sectors of society using technology and it is important to consider that the technology must be friendly with users, and even adapt to their needs and desires. We can see in current systems the user has to learn to use the devices or services, but with an intelligence system, the technology is a very effective way to determine the needs of users. In this research, we present the physical implementation of a system to assist users and patients in daily activities or duties. The system include a architecture of agents where a Deliberative agent learns from the interaction with the user, in this way the system detects thermal comfort preferences for give an automatic assistance. We propose an algorithm with a proactive stage and learning stage adapting a classification algorithm. We select the classification algorithm with the best performance using cross validation. The algorithms of pattern recognition was Back Propagation neural network, Naive Bayes, Minimum Distance and KNN (k near neighbor). Our motivation of this work was to help people with motor difficulties or people who use wheelchairs, for this reason it was essential to use a wireless controller and use a friendly interface. The system was implemented in a testbed at the Leon Institute of Technology in Guanajuato, Mexico, and include sensors of humidity and temperature, windows actuators, wireless agents and other devices. Experimental tests were performed with data collected during a time period and using use cases. The results were satisfactory because it was not only possible remote assistance by the user but it was possible to obtain user information to learn comfort preferences using vector features proposed and selecting the classification algorithm with better performance.