ADISTES
Active diagnosis aims at significantly improving system reliability by using diagnostic information at run-time for fault isolation and online error recovery. Active diagnosis for open embedded real-time systems (e.g., health management and medical systems) is an open research problem due to stringent real-time and reliability requirements in combination with constituent components that are unknown at design time. The ADISTES project extended semantic techniques, usually used in large-scale IT systems, for active diagnosis in open embedded real-time systems. We developed modeling techniques for expressing diagnostic features, symptoms, faults and recovery actions. Methods for distributed knowledge management established relaxed consistency while ensuring real-time constraints. Real-time inference was investigated based on the time-triggered scheduling of diagnostic queries. The goal of query transformations, semantic transformations and goal-oriented learning improved schedulability and reliability. The methods and algorithms were prototypically implemented, as well as experimentally and analytically evaluated concerning reliability and timeliness.
Description
In recent years, the field of embedded systems has evolved towards novel application areas that combine stringent real-time constraints, reliability requirements and the need for an open-world assumption. These systems are called open embedded real-time systems. Examples are Ambient-Assisted Living (AAL) systems for elderly care, networked medical devices and health management systems, applications for electrical power distribution and command/control systems. These systems are based on an open-world assumption where new components are integrated at run-time in order to dynamically realize emerging global services. At the same time, reliable operation and support for stringent real-time requirements are essential to support closed-loop control and guaranteed response times. For example, physicians need to dynamically integrate medical devices into an in-home AAL system for emergency treatment, while ensuring predict-able response times and reliable interaction with in-home devices (e.g., medical sensors) and remote sites (e.g., hospi-tal).
In this context, the research project ADISTES introduced models and algorithms for active diagnosis in open embedded real-time systems with improved reliability and safety. The project results enable cost-effective fault isolation and re-covery actions at run-time based on the time-triggered execution of diagnostic queries. In contrast to other fault-toler-ance techniques based on spatial and temporal redundancy, the incurred overhead is significantly reduced. At the same time, we cope with dynamic system structures, where components enter and leave at run-time, interact among each other in variable setups and realize different global application services. In this context, diagnostic relationships cannot be expressed absolutely, e.g., by defining assertions on specific state variables. Diagnostic relationships need to be mod-elled indirectly by referring to semantic categories and recurring diagnostic patterns. Systems with a dynamic structure must be observed and analysed upon the occurrence of anomalous behaviours and states. Likewise, the definition and execution of recovery actions must consider the dynamic nature of an open system.
The ADISTES project extended semantic techniques, usually used in large-scale IT systems, for active diagnosis in open embedded real-time systems. Based on this we developed novel modelling techniques for expressing diagnostic fea-tures, symptoms, faults, and recovery actions. Methods for the management of the semantic knowledge with a relaxed consistency and methods for automatically identifying and modelling faults were developed. Connecting the semantic and real-time environment, a diagnostic multi-query graph (DMG) enables to track and conclude on faults, which may lead to failures. The resulting real-time inference is established based on the time-triggered scheduling and optimization of diagnostic queries. The methods and algorithms were prototypically implemented, as well as experimentally and analytically evaluated concerning reliability and timeliness.
The ADISTES results have resulted in four PhD theses, several master theses, and numerous publications at scientific conferences. The ADISTES results are also leading to follow-up projects with companies in different domains such as automotive diagnosis and medical services funded by KMU Innovativ. A project in the automotive domain serves for the maintenance-oriented diagnosis for garages, while the medical projects aims at a wearable computing device for im-proved medical diagnosis of ambulance teams.
Focus points/areas
-
Active Diagnosis based on Semantic Web Technologies for Distributed Embedded Real-Time Systems
Methodology
The project consists of four Work Packages (WPs).
WP1: Finalization of the gap identification process in Iranian and Iraqi HE curricula is based on demands and needed skills in the private sector and industry as well as state-of-the-art technological advances in Internet of Things.
WP2 addresses the management of the DKB and self-learning.
WP3 provides scheduling and graph transformation algorithms for rule-based inference.
WP4 performs the experimental and analytical evaluation.
