The transition towards 6G enables new architectural paradigms integrating communication and computation to support data-intensive vertical applications. In the mobility domain, vehicles are increasingly acting as mobile sensors, allowing to maintain digital twins of urban infrastructure and another avenue for monetizing vehicle data. However, realizing this in practice is challenging due to limited on-board computing resources, high data transmission costs, and regulatory compliance constraints. We present a network-augmented edge architecture for privacy-preserving processing of vehicle sensor data, aligned with emerging 6G architectural principles. Our approach combines lightweight on-board pre-processing with confidential edge computing in the mobile network. We leverage confidential computing to ensure security and privacy of workloads while augmenting the on-board vehicle compute capacity. We practically demonstrate a use case where vehicles collect sensor data in a live urban setting and transmit selected data points to confidential edge nodes updating the digital representation of road signage. To ensure regulatory compliance, the approach was assessed as part of a personal data protection regulatory sandbox. In this paper, we discuss design choices, practical deployment aspects, and privacy considerations to support such capabilities in upcoming mobile networks