How are AI models evolving beyond object detection to predict and adapt to the behaviour of pedestrians, cyclists, and other vehicles in real time?
What are the key hardware and software breakthroughs needed to overcome compute limitations, sensor fusion challenges, and integration roadblocks in scaling L2+ ADAS to higher autonomy?
How far can real-time connectivity push autonomous decision-making, and what are the practical limitations of V2X, cloud-based learning, and cooperative manoeuvres in dense urban environments?
What are the most effective redundancy strategies in sensors, compute, and control logic to ensure fail-operational safety and prevent critical system failures?
How can large-scale simulation, fleet data collection, and AI-driven edge case analysis accelerate validation and real-world deployment of ADAS features?