According to XPENG (@XPengMotors) on Twitter, CEO He Xiaopeng conducted an in-depth test of nine real-world driving scenarios to evaluate the progress of their AI-powered autonomous driving system, the 2nd Generation VLA. The test used identical routes and rules in authentic traffic conditions. Results showed the VLA system delivered smooth rides, exhibited human-like reaction patterns, and avoided harsh robotic braking, achieving 100% confidence across all scenarios. This demonstration highlights tangible advancements in deep learning-based perception and decision-making for urban mobility, and positions XPENG's AI as a competitive player in the autonomous vehicle market. The business implication is clear: enhanced AI performance in complex environments could accelerate commercial adoption of smart driving features, improve user experience, and strengthen XPENG’s market position in China and globally (Source: XPENG Twitter, Jan 3, 2026).