跳转到主要内容

自动驾驶数据闭环专家

Technology
北京掌上先机网络科技有限公司
上海市, 中国¥30,000 - ¥60,000 /月1周前
全职

职位描述

该职位来源于猎聘 Role Overview Scania China is rapidly shifting from rule-based logic to an entirely data-driven, End-to-End (E2E) autonomous driving paradigm for heavy-duty vehicles. We are seeking a visionary Data Loop Architect to design and scale our vehicle-to-cloud infrastructure. You will own the architecture of Scania China's "Data Engine"—responsible for intelligent data ingestion, automated MLOps training, and generative simulation loops, while ensuring strict data compliance within the Chinese market.

Key Responsibilities

  • Vehicle-to-Cloud Ingestion: Design low-latency data pipelines and edge-side "Shadow Mode" triggers to automatically capture and backhaul critical disengagements and long-tail corner cases from test fleets.
  • Cloud Data Factory: Build highly automated data-processing pipelines leveraging foundation models for 4D reconstruction and natural language scene-tagging into a searchable Scene Bank.
  • Automated MLOps: Establish a fully automated training loop that seamlessly triggers data augmentation, model fine-tuning, and regression testing as soon as new data enters the system.
  • Simulation &

    • Validation: Architect the bridge between real-world driving data and generative AI world models to enable closed-loop, reactive simulation testing.

  • Data Sovereignty &

    • Compliance: Ensure all data collection, processing, and storage workflows strictly comply with China’s automotive data security regulations and surveying laws. Requirements

  • Education: Master’s degree or equivalent in Computer Science, Data Engineering, Robotics, or a related field.
  • Experience: 5 years in autonomous driving engineering, MLOps, or large-scale cloud data architecture.
  • Data &

    • AI Tech Stack: Proven hands-on experience with Big Data tools (Kafka, Spark, Flink), Deep Learning frameworks (PyTorch), and MLOps pipelines (Kubeflow, MLflow).

  • Automotive Know-How: Solid understanding of modern vehicle E/E architectures, Automotive Ethernet, and CAN data-logging protocols.
  • Market &

    • Language: Deep familiarity with Chinese data compliance regulations. Professional fluency in both English and Mandarin to effectively connect Scania China with Global HQ in Sweden.

    Preferred

    Qualifications

  • Demonstrated experience launching a data loop or shadow mode system from initial concept to mass production (SOP). Familiarity with commercial vehicle dynamics and specific heavy-duty truck use cases.

Keywords
PyTorchApache KafkaApache SparkCloud computingKubeflowDeep learningBig data

对这个职位感兴趣吗?