Qcraft provides a China‑first large‑scale AI model that jointly fuses multi‑modality sensor data and temporal information to deliver high‑precision perception and veteran‑driver‑level planning and control for autonomous driving. Its end‑to‑end navigation‑on‑autopilot solution supports point‑to‑point assisted driving, parking assistance, and active safety across vehicle platforms, while a closed‑loop data workflow streamlines labeling, training, simulation, and validation to accelerate development and reduce costs.
Funding
$100M raised to dateRaised to date based on public sources. This may differ from the amount the company actually raised and is based only on what is publicly available on the internet.
Founders
Product
Problem
Current autonomous driving systems struggle with fragmented sensor data processing, limited adaptability across diverse traffic scenarios, and inefficient development cycles that hinder rapid deployment of reliable advanced driver assistance and self‑driving solutions.
Solution
Qcraft delivers a China‑first large‑scale AI model that jointly fuses multi‑modality sensor inputs and temporal information on a production‑ready computing platform. Its spatio‑temporal algorithms generate flexible, veteran‑driver‑level planning and control strategies for both urban and highway environments. The company offers end‑to‑end navigation‑on‑autopilot (NOA) solutions that cover point‑to‑point assisted driving, parking assistance, and active safety across multiple vehicle platforms. By streamlining the data pipeline—from selection and labeling through training, simulation, and closed‑loop validation—Qcraft accelerates the development and verification of autonomous driving functions while reducing cost and complexity.
Target Audience
Primary customers are automotive manufacturers, tier‑1 suppliers, and fleet operators seeking scalable advanced driver assistance or autonomous driving capabilities for urban and highway vehicles.
Features
- Joint multi‑modality and temporal fusion large model enabling high‑precision perception across cameras, LiDAR, and radar
- Spatio‑temporal planning algorithm that produces adaptive, veteran‑driver‑like driving strategies for complex city and highway scenarios
- End‑to‑end city NOA solution achieving 6 m lane‑keeping accuracy on a single trip, supporting point‑to‑point assisted driving, parking, and active safety
- Modular sensor configuration options to suit different vehicle architectures and cost targets
- Closed‑loop data workflow integrating data selection, labeling, training, large‑scale simulation, and validation for rapid iteration
- Product families (e.g., “轻舟” series) offering customizable autonomous mobility spaces, from driverless shuttles to smart connected buses