AI in Bridge and Road Infrastructure Inspection Quiz

Answers and explanations appear as you take the quiz.

1. Under the US National Bridge Inspection Standards, how often do most highway bridges typically receive a routine inspection?
   • Every 6 months
   • Every 5 years
   • Every 24 months
   • Only after an accident
2. Why is crack detection usually treated as a segmentation task instead of simple object detection?
   • Practitioners need crack length and width, which requires pixel-level outlines
   • Segmentation needs no training data
   • Object detection cannot run on drones
   • Cracks are always perfectly rectangular
3. What is ground sampling distance?
   • The distance a drone flies between batteries
   • The depth ground-penetrating radar can reach
   • The spacing between bridge piers
   • The real-world size represented by one image pixel
4. Which tools does the guide name for revealing deterioration inside concrete decks?
   • Dashcams and GPS
   • Ground-penetrating radar and infrared thermography
   • Laser profilers and roughness indexes
   • Strain gauges only
5. Which of these is a common source of false positives for crack-detection models?
   • Exposed rebar
   • Potholes
   • Stains, joints, shadows and formwork marks
   • Corrosion on steel
6. In this guide, what is a digital twin of a bridge?
   • A model of the asset linked to its inspection history and sensor data
   • A second bridge built next to the first
   • A backup copy of the inspector's notes
   • A drone that follows the inspector
7. Why can't a crack image alone determine how serious the crack is?
   • Cameras cannot capture cracks
   • All cracks are equally dangerous
   • Images are too large to analyze
   • Seriousness depends on the crack's location, the type of member and the loads it carries
8. Which technique helps with the severe class imbalance in crack segmentation?
   • Using only color images
   • Dice or focal loss functions
   • Flying the drone higher
   • Removing all images that contain cracks