AI in Manufacturing and Industrial Processes AIE Practice Questions (Page 5)

Questions 21–25

1. 21expert · hard

   A plant has a critical compressor with sensors streaming data to a cloud platform. The AI team has built a predictive maintenance model that achieves 95% accuracy in predicting failures. However, the maintenance team complains that the model generates too many false alarms, causing them to ignore alerts. The plant manager wants to reduce false alarms without significantly increasing missed failures. Which action is most appropriate?

   • AAdjust the model's decision threshold to require a higher predicted probability before triggering an alarm, and validate the new threshold against historical failure data.
   • BRetrain the model with more historical data, assuming that more data always improves precision.
   • CReplace the predictive model with a simpler rule-based system that alarms only when vibration exceeds a fixed limit.
   • DIncrease the frequency of sensor data collection to get more granular data for the model.
   Reveal Solution Discussion Select an answer first
2. 22foundation · easy

   Which type of data is most commonly used by AI models for predictive maintenance in industrial equipment?

   • ASensor data such as vibration, temperature, and pressure
   • BCustomer feedback surveys
   • CFinancial transaction logs
   • DEmployee attendance records
   Reveal Solution Discussion Select an answer first
4. 23foundation · easy

   In inventory management, what is a key benefit of using AI?

   • AMaintaining optimal stock levels to reduce holding costs while avoiding stockouts
   • BGuaranteeing that inventory is always zero to save space
   • CAutomatically rejecting all supplier deliveries that are late
   • DEliminating the need for safety stock entirely
   Reveal Solution Discussion Select an answer first
5. 24expert · hard

   A manufacturer produces metal parts with two types of defects: surface scratches and internal cracks. The current vision system detects surface scratches well but cannot detect internal cracks. The company wants a single automated inspection solution that detects both defect types without slowing the production line. Which approach is most appropriate?

   • AImplement a digital twin that simulates the manufacturing process and predicts where defects are likely to occur, eliminating the need for physical inspection.
   • BUse a robotic arm to physically scan each part with a laser profilometer to detect both surface and internal defects.
   • CUse a combination of high-resolution optical cameras for surface inspection and ultrasonic or X-ray sensors for internal inspection, with AI models analyzing both data streams.
   • DTrain a single computer vision model on optical images to detect both surface and internal defects by analyzing shadows and reflections.
   Reveal Solution Discussion Select an answer first
6. 25foundation · easy

   How does AI-powered computer vision improve quality control in manufacturing?

   • ABy using cameras and image recognition to automatically detect defects in products
   • BBy manually reviewing product samples to ensure they meet standards
   • CBy analyzing audio signals from machines to detect quality issues
   • DBy replacing all physical inspection tools with laser measurement devices
   Reveal Solution Discussion Select an answer first