HASEM Spatial Dataset: Predictive Take-Over Request (TOR) Decision Zones and Centroids for High-Capacity Transport on Arctic Roads

Description This dataset contains the spatial outputs of the Highway Anomaly Segmentation Engine Model (HASEM). It includes the high-risk kinematic anomaly zones and the final geometric centroids used to define predictive Take-Over Request (TOR) decision nodes along the E6 highway corridor in northern Norway. File Structure The dataset consists of two spatial vector files in GeoJSON format (EPSG:4326): Top200_Zones.geojson: Spatial representation of the 200 most critical highway segments (150 m each) exhibiting the highest kinematic instability. These zones were identified using the maximum Kinematic Anomaly Index (Cs )derived from continuous heavy-duty vehicle operations. Clusters_Centroids_Geometric.geojson: The final 10 predictive TOR decision nodes. These geometric centroids were generated via K-Means clustering (k=10) of the top 200 risk zones and represent the exact spatial coordinates for preemptive SAE Level 3 system deactivation. Methodological & Confidentiality Note To protect commercial confidentiality and the operational security of the transport fleet, the raw event-driven GNSS telemetry logs (including exact speeds and timestamps) used to calculate the (Cs) index are not included in this public repository. Instead, this dataset provides the final, anonymised spatial aggregates. These files are intended to support the geographical reproducibility of the HASEM spatial mapping. Relation to the accompanying manuscript This spatial dataset supports the geographical reproducibility of the HASEM methodology and the location-based mapping of the TOR decision nodes detailed in the accompanying research article. (Keywords) HASEM; Spatial Analysis; High-Capacity Transport; SAE Level 3; Take-Over Request; Arctic roads; Geofencing; GeoJSON