Space-filling curves (SFCs) enable computationally efficient event look-ups by aggregating multi-dimensional data to their corresponding single dimensional representation. We have pioneered computationally efficient event identification for data in the automotive context (for example, pedestrian crossing identification) in research projects funded by Vinnova and VR. So far, we have explored the use of SFCs for structured vehicle motion signals to validate the applicability of our approach. Next, we want to extend the theories around feature extraction and SFC-based dimensionality aggregation to unstructured data from automotive context such as video data. Datasets relevant for our research community include Waymo, Zenseact, nuScenes, SMIRK, comma.ai, PIE, JAAD, and PREPER that cover different traffic situations from various geographical regions. Our current research is focusing on extending the theoretical applicability of SFCs for unstructured data streams and to systematically analyze their performance when scaling datasets. This proposal is linked to the following two ongoing projects: NAISS 2025/23-40 NAISS 2024/22-1517