AI case study

Li AutoDriving simulation

Reconstruction cost capped how much fleet data Li Auto could simulate. A unified camera-LiDAR model halved cost and doubled scene coverage.

Published

Key results

Reconstruction Cost Reduction
~50%
Training Speedup
4x
Scene Coverage Increase
2x

Result highlights

Unlock 3 result highlights

The story

Context

One of China's leading intelligent electric vehicle manufacturers, with a fully self-developed ADAS platform built around a closed-loop simulation environment where per-scene reconstruction cost sets a direct ceiling on how much fleet driving data can be used for algorithm validation.

Challenge

Separate camera and LiDAR reconstruction pipelines increased engineering overhead and made cross-sensor simulation consistency hard to maintain. Long...

Solution
Unlock full story

Quotes

The company

Li Auto logo

Li Auto

liauto.com

Manufacturer of premium smart electric vehicles.

IndustryAutomotive & Mobility
LocationBeijing, China
Employees10K-50K
Founded2015

The vendor

NVIDIA is a technology company that specializes in semiconductors, graphics processing units, and artificial intelligence for applications in data centers, gaming, and more.

IndustryTechnology
LocationSanta Clara, California, United States
Employees10K-50K
Founded1993

Use case

Li Auto's Driving simulation is part of this use case:

Autonomous Systems
29 case studies(+50% YoY)
Proven impact?
Average
3.9Typical
3.9Typicalwithin Product Engineering

Similar Case Studies

Related implementations across industries and use cases

29 AI case studies in Autonomous Systems

76 AI case studies in Automotive & Mobility

757 AI case studies in Product Engineering