The researchers involved in the Transfer of Knowledge (ToK) programme between Industry and Academy have the following objectives.

The three scientific objectives are:

1)     Studying the dependability of OTS based SCSs through dependability measures (e.g. Mean Time To Failure, Time To Failure, and statistical characterization of Failure Occurrences) and risk assessment (e.g. using software fault injection) both at the component and system level once OTS candidates are integrated. Different evaluation techniques will be investigated in order to determine the most suitable one for a specified class of OTS components (i.e. operating systems, virtual machines, and middleware infrastructures). The project will draw-up guidelines to support software engineers during the OTS selection phase (showing where and how proper evaluation tools have to used) and during the OTS integration phase (where and how system evaluation tools are needed).

2)     Studying and evaluate robustness of OTS components, i.e. how they behave under faulty conditions, and how they interact with other components in the target execution environment (Air Traffic Management Systems, complex industrial plants, etc.). Different fault-injection techniques will be applied to generate test cases (e.g., malformed invocations, combination of invalid input data types) in order to measure the robustness failure rate, i.e. the percentage of non-properly handled erroneous inputs and develop a prototype robustness assessment tool.

3)     Using the results of the previous analyses, to study and develop techniques for on-line monitoring and on-line diagnosis of fault occurrence in the operational phase of a SCS. On-line fault diagnosis is the process of determining the cause of errors, both in location and in nature, during system execution through fault detection (triggering of an alarm) and isolation (identification of error’s root cause and its propagation pattern). CRITICAL STEP diagnosis mechanisms will focus on the valuation of the following quantitative parameters: latency (i.e., the time  required to identify the root cause of a detected fault), accuracy (i.e., the probability that the occurred fault f is recognized by the diagnostic mechanism),  credibility (i.e., the probability that the diagnostic mechanism marks as faulty a component which is the real root cause of the fault), and coverage (i.e., the ratio between the number of recognized faults and the cardinality of the fault model set).

This joint research will profit from the synergetic effects of working in inter-sectorial teams, where academic researchers will better understand industrial constraints and “real-life” SCSs requirements in order to build up by the end of the project the scientific framework necessary to develop integrated methods and commercial tools for creating and handle the next generation Safety-Critical Systems, having a measurable safety or being endowed with measurable safety properties.

The three long-term strategic business objectives and challenges are to implement these know-how and breakthroughs into profitable products and services:

1)     Developing industrial strategies and marketable tools (e.g. fault-injection and quantitative analyses) for evaluating the robustness and dependability level of OTS components in SCSs, and thus making more effective the processes of OTS selection/integration into a complex system;

2)     Realising appropriate techniques/mechanisms and tools for on-line monitoring, diagnosis and dynamic reconfiguration of SCSs and thus assuring a defined and standardised safety level during the SCSs’ operational phase;

3)     Devising highly competitive, dependable, robust and certifiable safety critical software systems for both public and private end-users in Europe as well as other continents, outperforming foreign software and system developers in terms of quality, based on an open source middleware platform for Mission Critical and Near Real-Time applications (as for example, in Air Traffic Management).