Industrial Requirements For Supporting Ai Enhanced Model Driven Engineering

Industrial Requirements for Supporting AI-Enhanced Model-Driven Engineering by J Bergelin and P E Strandberg will be published soon at the Workshop on Artificial Intelligence and Model-driven Engineering here: [1] In short, it covers high-level generic requirements from the AIDOaRt project that covers AI, model-driven engineering (MDE), and DevOps - we put a special emphasis on MDE.

Pre-print at arXiv [2]

Tentative Abstract
There is an increasing interest in research on the combination of AI techniques and methods with MDE. However, there is a gap between AI and MDE practices, as well as between researchers and practitioners. This paper tackles this gap by reporting on industrial requirements in this field. In the AIDOaRt research project, practitioners and researchers collaborate on AI-augmented automation supporting modeling, coding, testing, monitoring, and continuous development in cyber-physical systems. Through a process of elicitation and refinement, 78 high-level requirements were defined, and generalized into 30 generic requirements by the AIDOaRt partners. The main contribution of this paper is the set of generic requirements from the project for enhancing development of cyber-physical systems with artificial intelligence, DevOps, and model-driven engineering, identifying the hot spots of industry needs in the interactions of MDE and AI. Future work will refine, implement and evaluate solutions towards these requirements.

Generic Requirements
Requirements Engineering

• RE.1, The system shall translate requirements from semi-structured language to formal language.
• RE.2, The system shall refine requirements expressed in formal language from analysis results.
• RE.3, The system shall produce suggestions or prescriptions for the requirements or systems engineer.
• RE.4, The system shall verify consistency of requirements from analysis results.

Modeling

• Mod.01, The system shall verify high-level models with AI techniques.
• Mod.02, The system shall generate tests using formal models, automated reasoning and/or AI methods.
• Mod.03, The system shall identify system failures using AI-based models.
• Mod.04, The system shall verify models with semi-automatic model synthesis.
• Mod.05, The system shall abstract models with AI-based approaches.
• Mod.06, The system shall provide easy configuration via AI-based methods.
• Mod.07, The system shall represent AI with a Model-based approach.
• Mod.08, The system shall perform retrospective analysis of model quality using AI-augmented methods.
• Mod.09, The system shall extend standards modeling languages to support aspect-oriented system development.
• Mod.10, The system shall deploy a solution on the cloud as a set of corresponding services with execution traceability.
• Mod.11, The system shall integrate DevOps workflows with CI/configuration models.

Coding

• Code.1, The system shall write abstractions of implementation with AI methods.
• Code.2, The system shall write refinement of specification with AI methods.
• Code.3, The system shall import/export variables/procedures between code and models.

Testing

• Test.1, The system shall generate test cases for high-level models with AI techniques.
• Test.2, The system shall perform automatic execution of test cases.
• Test.3, The system shall perform automatic verification of system architectures.
• Test.4, The system shall perform automated evaluation of testing results.
• Test.5, The system shall report or fix the problems detected in the testing phase.
• Test.6, The system shall integrate and analyze the testing phase into the DevOps pipeline.

Monitoring

• Mon.1, The system shall access different artifacts.
• Mon.1.1, The system shall access off-line data.
• Mon.1.2, The system shall access on-line data.
• Mon.1.3, The system shall access system logs and execution traces.
• Mon.1.4, The system shall access system states such as resource usages during monitoring.
• Mon.1.5, The system shall access traffic between nodes.
• Mon.1.6, The system shall access hardware input/output.
• Mon.1.7, The system shall access Documents expressed in natural language.
• Mon.2, The system shall support monitoring for many purposes.
• Mon.2.1, The system shall monitor with the purpose of identifying violations to pre-defined requirements.
• Mon.2.2, The system shall monitor for the purpose of identifying deviations, anomalies or security events.
• Mon.2.3, The system shall monitor for the purpose of identifying clusters of anomalies of some sense.
• Mon.2.4, The system shall monitor for the purpose of identifying (unwanted) trends or drifts.
• Mon.2.5, The system shall monitor for the purpose of identifying deviations between a system and a model.
• Mon.2.6, The system shall monitor for the purpose of identifying root-causes to anomalies/deviations.
• Mon.2.7, The system shall monitor for the purpose of minimizing downtime or other inefficiencies/waste.
• Mon.2.8, The system shall provide predictions of future (unwanted) behaviours based on observed features.
• Mon.2.9, The system shall monitor for the purpose of tracking system usage or user activities.
• Mon.3, The system shall discriminate between levels of automation during monitoring.
• Mon.4, The system shall produce artifacts for re-use and further training.
• Mon.5, The system shall provide human-in-the-loop functionality supporting heterogeneous data sources.
• Mon.6, The system shall motivate its actions during monitoring.

Tillhör Kategori Publikationer