Artificial intelligence technologies are improving user experience in smart vehicles, according to findings by J.D. Power and Shanghai-based Tongji University.

Their research, of which the findings were released on Thursday, show that such technologies as large language models have improved vehicles' interactive capabilities, enabling functions such as voice recognition, emotion analysis and personalized recommendations.

These advancements deliver an intelligent experience to vehicle users, they said.

The research shows the success rate of voice command stands at 98.6 percent for commonly used functions in top-rated models, meeting daily user demands.

However, the average number of steps to complete tasks on the central console screen has increased to 2.42, up by 0.56 steps from 2023.

This indicates a decline in interaction efficiency caused by rising system complexity and feature stacking.

The research also shows that the intelligent cabin features, page layouts and visual styles lack differentiation between brands and models. Additionally, some of the features lack real-world scenarios.

Automakers are encouraged to prioritize real user scenarios and integrate their brand values to create differentiated experiences.

"Whether automakers can reasonably reduce overly designed intelligent features while meeting user needs will be critical for the future of intelligent cabins," said Su Jun, president of J.D. Power China.

"Intelligent design should not simply pile up features but instead focus on practical user needs, using reasonable configurations to deliver better experience designs. This will be a key area for exploration in the coming years," said Su.