Building the AI-native data infrastructure for intelligent connected vehicles
With the deep integration of AI technology and the automotive industry, EMQX Platform redefines the data value loop for intelligent connected vehicles through its innovative data infrastructure capabilities.
In-Vehicle and V2X Communications with AI-Optimized MQTT Platform for Next-Gen Automotive Intelligence
We offer the automotive industry a modern, AI-optimized MQTT platform. We pioneer MQTT over QUIC, the next-gen standard protocol for IoV messaging, serving over 50 automotive companies and connecting more than 10 million electric and traditional vehicles.
Capabilities
Bi-Directional Communication
Supports bi-directional communication, allowing vehicles to send telemetry data and receive commands or updates from the cloud.
Real-Time Data Transmission
Guarantees vehicles and your IoV applications to deliver data in real-time with a single-digit millisecond latency.
MQTT over QUIC
Experience MQTT over QUIC, the next-gen standard protocol for IoV applications, optimized for vehicle mobility and unstable network connectivity.
File Transfer over MQTT
Streamline your IoV system design by consolidating data communication and file transfer into a single, efficient MQTT channel.
On-the-fly Data Transformation
Extract, filter, enrich, and transform in-flight data using an SQL-based rules engine in vehicles and in the cloud.
Stream Processing and AI/ML
Execute AI/ML modeling and streaming data processing within vehicles and cloud platforms for business optimization and intelligence.
Why EMQ
A vehicle-cloud collaborative data closed-loop architecture optimized for AI/LLM scenarios, meeting the evolving data demands of intelligent vehicles.
In-Vehicle Multi-Domain Heterogeneous Data Ingestion
- CAN bus， DDS, and SOMEIP integration， cross-domain bus data ingestion
- Support for multiple data types, including signals, files, and video
Full-Spectrum High-Precision Real-Time Acquisition
- Millisecond-level real-time acquisition of tens of thousands of in-vehicle signals
- Lightweight and Efficient Onboard Full-Data Compression & Storage
- Optimized compression and disk storage for onboard data
Onboard Edge Computing
- Real-time data parsing, filtering, downsampling, screening, and transformation onboard
- Custom metric computation and event detection
- Integration of algorithm models with real-time inference capabilities
Reliable Mass Connectivity
- Supports tens of millions of concurrent vehicle data connections
- Utilizes innovative MQTT over QUIC for reliable data transmission in weak network conditions
- Provides rich data sources for AI applications
Intelligent Data Processing
- Built-in rule engine and stream processing capabilities
- Real-time data filtering and transformation to enhance AI model training efficiency
AI Integration Layer
- Native support for Kafka, Snowflake, and other big data services
- Seamless integration with LLM training platforms, simplifying AI data pipelines
Use Cases
Accelerating Autonomous Driving Model Training
- Data Integrity Assurance: Automatic network recovery ensures continuous sensor data (LiDAR, cameras, radar) collection, preventing training data gaps.
- Efficient Data Supply: Billion-device cluster capacity meets large-scale fleet data collection needs, accelerating model iteration.
- Global Data Collaboration: Cross-region Cluster Linking enables distributed training data aggregation, improving model generalization.
Enhanced Smart Cockpit Experience
- Real-Time Context Awareness: Millisecond-level latency ensures natural voice interactions.
- Personalized Services: Combines driver behavior analytics with LLMs to generate tailored recommendations (music, navigation, etc.).
- Multimodal Fusion: Unified ingestion and processing of audio, image, and text data.
Predictive Maintenance & Remote Diagnostics
- Real-Time Vehicle Health Monitoring: Analyzes millions of vehicle data points to detect faults preemptively.
- Root Cause Analysis: AI models correlate multi-system data to pinpoint issues.
- Knowledge Base Construction: Event history logging accumulates repair cases to continuously optimize diagnostic models.