Bài kiểm tra có liên kết hướng dẫn · Trung bình Cấp độ

Bài kiểm tra AI trong sản xuất dược phẩm

Kiểm tra mục đích PAT, xác thực mô hình, tính toàn vẹn dữ liệu, kiểm soát thay đổi, phát hành hàng loạt và giám sát chất lượng con người trong các nhà máy được quản lý.

Đường dẫn hướng dẫn liên quanAI trong sản xuất dược phẩm
Câu hỏi 1 của 8

What does a manufacturing AI alert establish?

Xem trước mọi câu hỏi8 câu hỏi

Câu trả lời và giải thích xuất hiện khi bạn làm bài kiểm tra.

  1. What does a manufacturing AI alert establish?
    • A signal that requires investigation under validated procedures
    • That the batch automatically fails quality
    • That the batch is released
    • That a sensor is always correct
  2. Which purpose does FDA describe for its PAT framework?
    • Marketing claims about a medicine
    • Process understanding and innovative development, manufacturing, and quality assurance
    • A replacement for current good manufacturing practice
    • Automatic release without controls
  3. Why validate across equipment and operating ranges?
    • All manufacturing equipment behaves identically
    • Validation changes the product formulation
    • A model cannot read sensor values
    • Process conditions may differ from historical training data
  4. What should change control cover for an AI system?
    • Only changes in office staffing
    • No changes once a model is deployed
    • Model, data, sensor, and process updates
    • Only changes to the user interface
  5. Can a model output release a batch automatically?
    • Not unless it is qualified and authorized for that specific purpose within the quality system
    • Yes, any prediction counts as a release test
    • Yes, if the score is high
    • No process measurements may ever be automated
  6. Which data feature supports auditability?
    • A screenshot alone
    • An unrecorded software update
    • Traceability from sensor measurement to model version and output
    • A handwritten summary without timestamps
  7. What should happen if the model is unavailable?
    • Create a replacement result manually without recording it
    • Use the defined validated fallback procedure
    • Pause quality review indefinitely
    • Release product based on the last prediction
  8. Who remains accountable for product quality decisions?
    • The model vendor alone
    • The sensor manufacturer alone
    • The algorithm’s training data
    • The manufacturer’s quality system and authorized personnel