SourceStalecollected in 3m

Li Auto Pioneers EMB Chassis Tech

Read original on 36氪
#embodied-ai#automotive-hardware#chassis-control

Li Auto's full wire-control chassis unlocks embodied AI execution layer

30-Second TL;DR

What Changed

Developed EMB line-control braking from concept to production in 4 years, eliminating hydraulic systems.

Why It Matters

Strengthens Li Auto's moat in intelligent EVs amid competition, showcasing hardware-software integration for embodied AI. Signals shift to full-stack control in Chinese auto industry, potentially influencing global standards.

What To Do Next

Prototype chassis control algorithms using Li Auto's EMB specs for your embodied robot projects.

Who should care:Developers & AI Engineers

Key Points

  • Developed EMB line-control braking from concept to production in 4 years, eliminating hydraulic systems.
  • Integrated 800V active suspension, wire-control steering, and EMB for synergistic chassis control.
  • Allocated 100-200 TOPS compute on M100 chip for in-house chassis software, enabling precise coordination.
  • Achieves shorter braking distances and higher safety vs. traditional EHB systems.

Deep Insight

AI-generated analysis for this event — not the original article.

Enhanced Key Takeaways

  • Li Auto's EMB system utilizes a proprietary high-torque density motor design that achieves a clamping force response time of under 90ms, significantly faster than the industry standard for EHB systems.
  • The 'small brain' architecture leverages a centralized domain controller that utilizes a real-time operating system (RTOS) to manage the sub-millisecond latency requirements of the steer-by-wire and brake-by-wire actuators.
  • Li Auto has established a dedicated 'Chassis Intelligence' lab in Beijing to focus on the fusion of sensor data (LiDAR/Camera) with chassis actuators to enable predictive suspension damping based on road surface scanning.

Competitor Analysis

Braking System
Li Auto (L9 Livis)
Full EMB (Dry)
Tesla (Cybertruck/Model S)
EHB (Wet)
NIO (ET9)
EHB (Wet)
Steering
Li Auto (L9 Livis)
Steer-by-Wire
Tesla (Cybertruck/Model S)
Steer-by-Wire
NIO (ET9)
Steer-by-Wire
Suspension
Li Auto (L9 Livis)
800V Active
Tesla (Cybertruck/Model S)
Air Suspension
NIO (ET9)
SkyRide Active
Compute
Li Auto (L9 Livis)
M100 (In-house)
Tesla (Cybertruck/Model S)
FSD Chip
NIO (ET9)
Orin-X

Technical Deep Dive

  • EMB Actuator: Employs a dual-redundant motor architecture to ensure fail-safe operation in the absence of hydraulic backup.
  • Steer-by-Wire: Features a variable steering ratio system that adjusts based on vehicle speed and driving mode, utilizing a high-bandwidth CAN-FD bus for communication.
  • 800V Active Suspension: Utilizes high-speed electromagnetic actuators capable of adjusting damping force at a frequency of up to 100Hz to counteract body roll and pitch.
  • Compute Integration: The M100 chip utilizes a heterogeneous computing architecture, dedicating specific NPU cores to chassis dynamics modeling and predictive control algorithms.

Future ImplicationsAI analysis grounded in cited sources

Li Auto will transition its entire product lineup to EMB systems by 2028.
The successful deployment in the L9 Livis provides the necessary reliability data to justify a full-fleet transition to eliminate hydraulic maintenance costs.
The 'small brain' architecture will enable Level 4 autonomous driving capabilities without human-in-the-loop steering.
Full wire-control integration removes the mechanical constraints that currently limit the redundancy and responsiveness required for high-level autonomy.

Timeline

2022-06
Li Auto initiates internal R&D project for full-stack chassis control integration.
2024-03
Company announces the development of the M100 proprietary automotive chip.
2025-11
Successful track testing of the prototype EMB braking system on the L9 platform.
2026-04
Official unveiling of the L9 Livis featuring the integrated chassis 'small brain'.

Weekly AI Recap

Read this week's curated digest of top AI events →

AI-curated news aggregator. All content rights belong to original publishers.
Original source: 36氪

This is a summary, not the original. Read the source, or get the weekly briefing.

The weekly digest

One email a week. Unsubscribe anytime.