AI-DAPT project kick-off meeting
Reconciling science-guided models & data-driven approaches for AI models optimisation
01-02-2024
Following a successful 2-day kick-off meeting held in Athens on Jan 23-24, the AI-DAPT project has just started its journey!
AI-DAPT is a 42-month Horizon Europe Project that will develop an AIOps framework to support and automate data and AI pipelines that involve science-guided models (e.g. first-principle, physics-based models or optimization algorithms) and ML models in loosely/tightly coupled interactions. The ambition is to introduce end-to-end automation and AI-based methods to support the design, execution, observability and continuous adaptation of data and AI pipelines.
The project involves a large number of academic and industrial partners, including a number of large-scale demonstrators to validate AI-DAPT result in several domains: for Health (non-invasive Glucose monitoring), Energy (residential demand response through smart heating), Cognitive Ergonomics (human-centered automation) and Manufacturing (predictive maintenance of production assets).
The project involves a large number of academic and industrial partners, including a number of large-scale demonstrators to validate AI-DAPT result in several domains: for Health (non-invasive Glucose monitoring), Energy (residential demand response through smart heating), Cognitive Ergonomics (human-centered automation) and Manufacturing (predictive maintenance of production assets).
Within AI-DAPT, along with project coordination, Athena Research Center leads the key activity to design, develop and deliver novel methods for the AI Pipelines that will reconcile science-guided models and AI/data-driven approaches, and are responsible for the training, explainability, evaluation, execution and optimization of AI models. Lead of the technical coordination of the project is Suite5 Data Intelligence Solutions, closely working with Athena Research Center in several topics related to data-centric and explainable AI technologies (check also XMANAI: making AI understandable).