Definition
Defect detection effectiveness of robotic inspection compared to human entry in confined spaces.
The Measurement Gap
What This Metric Measures
RHIER measures the ratio of defect detection effectiveness between robotic or drone inspection and human entry inspection in confined spaces: ballast tanks, cargo tanks, and chain lockers. Effectiveness is measured as critical defects found per unit of inspection scope, comparing the robot's systematic area coverage against the human inspector's experience-driven selective inspection pattern.
The Decision It Supports
Is the robot finding the defects that matter, or is it achieving high area coverage while missing the experience-driven hotspots that a human inspector would prioritize? A robot covering 95% of a ballast tank at uniform density may find fewer critical defects per hour than a human covering 60% who concentrates on bracket toes, horizontal stringers at the waterline, and areas near heating coils where 90% of critical defects occur.
Why Legacy Metrics Cannot See This
Legacy inspection metrics track area coverage percentage and total defect count. The robot reports 95% area coverage and 15 defects. The human reports 60% area coverage and 12 defects. By coverage metrics, the robot wins. By total defect count, the robot wins. But 10 of the human's 12 defects are severity-critical structural findings in known hotspot areas, while 11 of the robot's 15 defects are minor coating breakdown on clean side shell that poses no structural risk. RHIER normalizes for this by weighting defects by severity and normalizing by area, revealing that the human finds more critical defects per square meter inspected because experience directs attention to where defects actually occur.
Formula
Primary Formula
RHIER = (Critical_defects_robot / Area_robot) / (Critical_defects_human / Area_human)
Variable Definitions
- Critical_defects_robot = severity-weighted defects found by robotic inspection (count, weighted by severity classification)
- Area_robot = area inspected by robot (square meters)
- Critical_defects_human = severity-weighted defects found by human inspection (count, same severity weighting scheme)
- Area_human = area inspected by human (square meters)
Units:
Ratio. A value of 1.0 indicates equivalent effectiveness per unit area. Values above 1.0 mean the robot is more effective per square meter. Values below 1.0 mean the human is more effective per square meter, which is expected for experienced inspectors who concentrate on known problem areas.
Edge Cases
True matched comparisons require both methods on the same tank: possible during validation programs or when class requires both robotic and human inspection. Human inspectors self-select high-risk areas with higher critical defect density per area. Normalize for area risk profile if data on known hotspot locations is available. Minimum sample of 10 matched tank inspections for meaningful comparison.
Proposed formula, pending validation pilots.