An Evolvable Knowledge Graph Supporting a Hybrid Intelligence Autonomous Driving System
Anna Teern, Nada Elgendy, Pertti Seppänen, Tero Päivärinta · 2025
Route planning is influenced by various factors, with speed and distance often prioritized in map services. However, dynamic conditions such as road maintenance, weather, or accidents can significantly impact these routes. By leveraging hybrid intelligence (HI), the collaboration between human intuition and machine efficiency, we focus on supporting decision-making in autonomous and semi-autonomous driving contexts. This study explores integrating knowledge of the dynamic conditions into HI Autonomous Driving Systems (HI-ADS) within the 6G Visible Project. The findings demonstrate the potential of knowledge graphs (KGs) to enhance decision-making by integrating evolving data and ensuring adaptability to real-world driving conditions. Based on the design objectives of learning for evolvable KGs, concrete requirements for the HI-ADS KG are established.