AIDOaRt: AI-augmented Automation for DevOps, a Model-based Framework for Continuous Development in Cyber-Physical Systems

Abstract

HAL is a multi-disciplinary open access archive for the deposit and dissemination of scientific research documents, whether they are published or not. The documents may come from teaching and research institutions in France or abroad, or from public or private research centers. L'archive ouverte pluridisciplinaire HAL, est destinée au dépôt et à la diffusion de documents scientifiques de niveau recherche, publiés ou non, émanant des établissements d'enseignement et de recherche français ou étrangers, des laboratoires publics ou privés.

Explainable AI for Cyber-Physical Systems: Issues and Challenges

“…As the field of explainable artificial intelligence grows, making more models available, stakeholders will be more willing to implement it in their systems and thus gain any resulting improvements from XAI such as decreased production time, decreased system downtime, and increased profits [13]. The implementation of AI in industry can have great benefits such as allowing systems to make decisions and take action quickly, often acting automatically without the need for direct human control [57]. Such implementation requires AI to not only work as intended but also to be safe, trusted, ethical, and legal.…”

“…Such implementation requires AI to not only work as intended but also to be safe, trusted, ethical, and legal. To achieve this responsible design, AI must be explainable and accountable [57]. Stakeholders must be able to justify AI-recommended decisions; XAI provides a basis for this.…”

AI-augmented Model-Based Capabilities in the AIDOaRt Project

“…The project started up on 1st April 2021 and involves 32 industrial and academic partners from different European countries [1]. The AIDOaRt project consortium consists of three distinctly different categories of organizations with respect to interests and plans.…”

The Westermo network traffic data set

“…Both these data sets have led to progress in knowledge in the form of validated ideas, algorithms and tools. The Westermo test system performance data set was created during the AIDOaRt project [2] to explore anomaly detection. Figure 1 illustrates the load-15m performance metric over time for five test systems.…”