To develop automated and autonomous driving systems (ADS) it is important to understand the environment in which the system will operate. Operational Domains (OD) are defining the environmental conditions of service areas such as infrastructural or weather situations. Collecting relevant data in required scope (e.g., road-level vs. lane-level) has an impact on the possible Operational Design Domain (ODD) conditions and therefore has impact on the scope of potential limits of the operation, too. We want to show how data-driven discovery can support providing relevant data by transforming it into the right level of detail as Current Operational Domain (COD) and how AI supports creating proper ODD conditions based on relevant CODs.

• Which data and level of detail are relevant for ODD definition?
• How does data-driven ODD definition look like?
• How can AI support the data discovery and ODD definition and which boundaries exist?