Andrea Claudi

A goal-oriented requirements engineering approach for the ambient assisted living domain

The Ambient Assisted Living (AAL) domain is associated with a large number of stakeholders such as patients, their relatives, caregivers and physicians. This variety introduces a great heterogeneity in system requirements, which sometimes results in conflicting needs that must be considered when developing effective AAL systems. In this work we adopt a Goal-Oriented Requirements Engineering (GORE) approach to map out needs and requirements for the AAL domain. Following the requirements mapping, we also propose a preliminary architecture for a home care system (named e-Ward) to assist patients in their domestic environment as if they are in a hospital room.

Secure Data and Voice Transmission over GSM Voice Channel: Applications for Secure Communications

The GSM voice channel is the worlds most widely used mobile communication network. Unfortunately these networks are affected by serious vulnerability from hardware-based attacks and communications can be easy to intercept. This paper describes a prototype implementation of a device which enables secure data and voice transmission over GSM voice channel network. The proposed system is supposed to interact with every cell phone or mobile device connected to the GSM network, differentiating itself from the best-known crypto-phone category. In the following, we will highlight the particular modulation strategy used to generate a specific wave-form that allows the receiver to correctly decipher data with a very low BER. In conclusion, this work aims to define the principles to model a secure infrastructure over a GSM network.

MARVIN: Mobile Autonomous Robot for Video Surveillance Networks

The use of automated surveillance systems is common in several application domains, from transport and commerce to military. Surveillance systems evolved through the years: first and second generation systems use fixed or semi-mobile cameras, while third generation systems make use of autonomous robots. This way intelligent surveillance systems turn into distributed systems, and their configuration can vary to take into account particular environmental conditions or to compensate for one or more malfunctioning sensors. This paper describes MARVIN, a mobile autonomous robot to be used in a video surveillance system. Using well-established face-detection and face-recognition methods, MARVIN can track and follow human faces in the monitored environment. This paper proposes also an experimental evaluation of the real-time performances of the system, under conditions specified in a well-defined reference scenario.

Combining Artificial Intelligence and NetMedicine for Ambient Assisted Living: A Distributed BDI-based Expert System

World population is shifting towards older ages: according to recent estimates there will be 1.5 billion people over 65 years old in 2050. Local governments, international institutions, care organizations and industry are fostering the research community to find solutions to face the unprecedented challenges raised by population ageing. A combination of Artificial Intelligence and NetMedicine could be ideal to face these challenges: they provide the means to develop an intelligent system and simultaneously to distribute it over a network, allowing the communication over the internet, if needed. Hence, the authors present a Multi-Agent Architecture for Ambient Assisted Living AAL: it is the model for a system to manage a distributed sensor network composed by ambient and biometric sensors. The system should analyse data and pro-actively decide to trigger alarms if anomalies are detected. The authors tested the architecture implementing a prototypical Multi-Agent System MAS, based on Belief-Desire-Intention BDI paradigm: the Virtual Carer.

Testing Linux-based real-time systems: Lachesis

Testing is a key step in the software development cycle. Errors and bug fixing costs can significantly affect development costs without a full and comprehensive test on the system. Testing is even more important in real-time systems, because they are used in environments where a single failure could lead to serious damages. The first efforts to change the Linux kernel, introducing in it some real-time features, are dating back more than ten years ago. Nevertheless, up to now theres not a comprehensive test suite able to assesses the functionality and the conformance to the real-time operating systems standards for the Linux kernel and for the real-time nanokernels that rely on it. This paper discusses an ongoing work on Lachesis, an automated test suite for real-time Linux features. Lachesiss main goals are: to provide a standard and simple API to develop new tests; to provide a set of functional, regression, performance and stress test, either developing or porting them from other test suites; to design and experiment a series of build tests; to make the test suite extensible and portable. Up to now Lachesis is used to test Linux, PREEMPT RT, RTAI and Xenomai real-time features and performances. It provides some tests for SCHED DEADLINE patch, too. Lachesis is under active development: extensions to other real-time systems and more tests are planned to be added in the near future.

Towards Multi-Agent Health Information Systems

One of the key challenges in the healthcare sector is to adapt Health Information Systems to requirements coming from changing societies. In recent years, governments and international healthcare organizations defined a series of requirements for new generation Health Information Systems: they have to preserve past investments on legacy systems, but must also integrate new technologies, include the patient among their users, and ensure that clinical information are available at all times, even in places far from where information are physically stored. This paper proposes a multi agent-oriented architecture for Health Information Systems, which uses international standards for communication and management of clinical documents. The architecture tries to effectively model a generic healthcare organization, and aims at being easily extensible and adaptable to the particularities of specific healthcare systems. The authors present two experimental scenarios to test the proposed multi-agent health information system. In the first, they show how to model a specific use case, a radiology workflow, using agents and well-known standards; in the second one the authors demonstrate how a mobile application can use the services provided by the agents to support the medical staff in an emergency situation.

A Multi-Agent Expert System Simulation for Ambient Assisted Living: The Virtual Carer Experience

A shift in the distribution of population towards older ages is occurring in almost every country of the world and it is becoming a major problem in Europe, Japan and USA, due to the high dependency ratio of these countries. Ambient Assisted Living (AAL) and Ambient Intelligence (AmI) are polarising the attention of the research community, trying to improve the quality of life of elderly people promoting their autonomy, self-confidence and mobility. Artificial Intelligence (AI), developing intelligent and adaptable systems, can play a considerable role in AAL and AmI to cope with the changing needs that characterize the life of people with chronic diseases. To show a possible contribution of AI in these fields, this paper introduces the multi-agent architecture of an expert system for Ambient Assisted Living: the Virtual Carer (VC). Based on the Belief-Desire-Intention (BDI) paradigm, it models the behaviours of a human caregiver. The main goal of the Virtual Carer is to help an elderly patient in his daily activities, while his health conditions are monitored, in order to ensure his security. To show the main capabilities of the system, the paper describes some simulations of the proposed agency, highlighting the architecture of the reasoning component.

A noise-robust obstacle detection algorithm for mobile robots using active 3D sensors

Obstacle detection is one of the most important tasks for mobile robots moving along a plane, and it is critical to avoid damages, either to the robot or to human operators. In the past decades, several techniques were proposed for visual navigation of mobile robots, relying on different kind of sensors and algorithms: sonar sensors, laser stripes, and stereo vision are commonly used techniques. Even if these techniques are well-established and used in commercial robots, different and better sensors are now widespread, such as depth sensors. This work proposes an algorithm based on the use of an active 3D depth sensor for obstacle detection and avoidance. The algorithm, conceived to be used in embedded systems with low processing power, underwent several experiments and proved to be robust to Gaussian white noise.