The research focus ExploIT Dynamcis deals with the observation, modelling, analysis and design of the dynamic behaviour of information technology systems (IT systems) in their system environment. The focus area aims to comprehensively improve the evolvability of software-intensive systems to frequently changing environmental parameters.

The main tasks of the research area include fundamental work on the development and application of sound methods and techniques for the observation of relevant quality properties of dynamic systems, the modelling of changes in these quality properties during system operation and the modification and influencing of the system and system environment with the aim of maintaining or improving these quality properties.

ExploIT Dynamcis bundles practical and applied Computing Science competences for the provision and maintenance of high-quality information technology systems in areas affected by major changes. The aim is to provide domain-independent information technology concepts and methods to make system changes possible and controllable and at the same time to maintain or further develop tried-and-tested, high-quality system components.

The research focus comprises a consortium of scientists from Computing Science that fully covers the described spectrum of the topic and has the corresponding expertise to efficiently and usefully apply the developed domain-independent concepts and methods in concrete application domains.

The dynamic behaviour of overall systems (consisting of the system and its system environment) is understood as the time-varying behaviour of a system and/or its system environment. The behavioural changes of information technology systems that occur over time are based on many possible causes:

• Hardware components can fail due to ageing and wear and tear and thus no longer adequately fulfil their tasks.
• Software components can no longer support the tasks assigned to them due to changes in requirements.
• Software components are subject to permanent change due to necessary conversions, e.g. Google records more than 20 code changes per minute or a change of around 50% of the code base per month.
• Topologies of system environments can change due to the addition or migration of components, e.g. due to necessary load distribution or changed requirement profiles.
• Human system users can change the way they work with the information system, e.g. due to changes in exogenous work guidelines or usage contexts.

Dynamic system behaviour harbours risks and opportunities: Risks arise when systems or their system environment change so much due to the dynamics that system-relevant, important quality characteristics can no longer be guaranteed. On the other hand, opportunities also arise from system dynamics if changing quality criteria can be achieved better or more easily through targeted system customisation.

The quality properties considered here include security, trust and confidence building, maintainability, scalability, reliability, availability, probability of survival, performance, correctness, validity, interoperability and portability. These quality characteristics lead to the observation and optimisation of the quality objective of adaptability and evolvability, which summarises the characteristics of systems for efficient and effective reaction to internal and external changes.

ExploIT Dynamcis aims to develop general information technology methods and techniques for mastering system dynamics. Therefore, the systems analysed in the focus area are not restricted to individual domains. They cover the full range of hardware, software and embedded systems, centralised systems, distributed systems, federated systems and heterogeneous systems.

The ExploIT Dynamcis focus area examines the dynamics of information technologysystems (IT systems) in theirsystem environment. This overall system(system) must be

1. observe (observe),
2. analyse the changes detected in the system and the system environment (analyse),
3. decide on necessary adjustments to the observed overall system resulting from the analysis results (decide), and
4. make these adjustments to the information technology system or the system environment under consideration (actuate) (see figure "ExploIT Dynamics control loop").

ExploIT Dynamcis pursues a model-based approach that considers information technology systems, their system environment as well as observation aspects, analysis approaches, decision bases and system adaptations on the basis of corresponding models.

The scientific challenges of the focus area therefore relate to the development of an adaptive, integrated modelling, analysis, evaluation and transformation concept that encompasses all levels of abstraction of information systems and is to be tested in specific application scenarios. These application scenarios should be defined as comprehensively as possible across the participating research groups in order to enable long-term joint research cooperation, which can also result in larger research networks.

The permanent change of software systems results in a sequence of system versions(system). Each of these versions changes the previous one on the basis of the respective decisions(decide, actuate) resulting from observed changes(observe) to the system environment (usually changes to requirements or changes to underlying software and hardware platforms) and their interpretation in the analysis phase ( analyse).

As part of the ExploIT Dynamcis focus, approaches are to be developed for modelling and documenting these changes to the respective system versions. These modelling approaches are to be applied across all abstraction levels of software systems (including code, architecture and requirements). This makes it possible to clarify the evolution of software systems and to gain insights into structured software evolution and its effects on software quality based on comprehensibly documented system changes and their quality assessment.\footnote{This project is being carried out as part of a doctoral project in the Erasmus Mundus project "TARGET" in co-operation between the departments of "Information Systems I" and "Software Engineering"}. Another aspect when considering software changes is the consideration of changing interfaces (and thus subject to evolutionary mechanisms). These represent a concrete example of communication channels between components of a system, touching on all three levels of abstraction mentioned above and are thus a concrete area of application for a part of the research to be carried out here.

Software migration is concerned with the transfer of a software system to a new environment without changing its functionality. Particularly in the case of language migration (e.g. from COBOL to Java), no valid or empirically proven statements on the change in software quality as a result of migration are known.

As part of the ExploIT Dynamcis focus area, procedures for comparative evaluation(observe, analyse) of the quality of the source texts are to be determined on the basis of standardised representations of source code(system, usually using abstract syntax graphs) in order to derive improvements in direct language transformation and measurable quality improvement(decide, actuate) of the target code. This allows the quality of language migration procedures to be compared and the improvement in the quality of the target code (also through downstream reengineering measures) to be made verifiable.\footnote{The Software Engineering department is currently discussing a co-operation project in this context.}

Replication is an effective method for increasing the availability of data and services: availability indicates the probability with which operations on data can be successfully carried out or services successfully utilised at a given time. As the smooth functioning of our society is increasingly dependent on the availability of data and the ability to successfully utilise Computing Science services, for example in the financial sector (example: stock market data), in the state security sector (example: retrieval of vehicle owner information), in the business sector (example: account access), in the emergency sector (example: access to hazardous substance information) and in the transport sector (example: retrieval of traffic jam information), the targeted guarantee of application-specific, minimum availability of data and services plays an essential role.

In the context of replication, data (or services) are realised several times in the form of so-called replicas and managed appropriately according to a so-called replication strategy. The problem of identifying a suitable replication strategy that, on the one hand, produces the application-specific, minimum availability of data or services and, on the other hand, causes only minimal costs for data access or service retrieval -- despite great efforts in the past -- has not yet been satisfactorily solved.\footnote{The problem of identifying suitable data replication strategies from a given set of possible solutions is currently being worked on in the context of a wireless sensor network as part of a doctoral project in co-operation between the departments of "System Software and Distributed Systems" and "Information Systems II".