Construction is characterized by engineer-to-order (ETO) production, high variability, and fragmented information flows, which challenge the implementation of data-driven production planning systems. This study investigates how autonomous and semi-autonomous data collection technologies can support continuous, data-driven replanning in construction in line with advanced planning and scheduling (APS) principles. An embedded single case study was conducted on a large renovation and new-build project in Sweden, comprising three use cases involving an unmanned ground vehicle and a helmet-mounted 360° camera. The technologies were evaluated with respect to their ability to capture site data and support planning-related tasks such as progress monitoring, site utilization planning, and workplace safety inspections. The findings show that autonomous and semi-autonomous data collection is technically feasible and can enhance situational awareness and planning support; however, integration between data capture, interpretation, and planning systems remains limited. Most collected data require human mediation, and fully automated feedback loops for continuous planning are not yet achievable with off-the-shelf solutions. The study is relevant to production planning research by empirically examining how APS principles can be adapted to ETO construction contexts and by identifying key technological, organizational, and data-related constraints. The results indicate that near-term value lies in semi-automated workflows that augment human decision-making rather than fully autonomous planning, providing a foundation for more resource-efficient and adaptive production systems.