Pony.ai

Designing the Passenger Interface for Safe, Legible Autonomy.

For the 2018 World Artificial Intelligence Conference in Shanghai, our goal was to design a passenger-first in-vehicle that introduced riders to Pony.ai’s autonomous driving system in a way that felt reassuring, legible, and intuitive at speed.

At the time, the engineering team’s HMI was optimized for developers and vehicle operators, not everyday passengers. It exposed raw sensor data and system states in ways that were technically accurate but cognitively dense. Our challenge was to translate autonomy into human terms. We focused on situational awareness, confidence cues, and progressive disclosure, ensuring riders could understand what the vehicle “saw,” what it was doing, and why, without feeling overwhelmed.

The interface prioritized only the most meaningful environmental inputs like vehicles, pedestrians, lanes, and decision states using a calm, structured visual hierarchy. This transformed complex perception data into clear, predictive storytelling, helping passengers build trust in the system’s capabilities throughout the ride.

The result was a confidence-driven, human-centered autonomous experience that bridged deep AI systems with real-world passenger comfort and understanding—turning advanced autonomy into something people could immediately feel safe with.