AUTOMATED INSPECTION OF AIRCRAFT
INTRODUCTION establish technical feasibility of using robotic systems in aircraft maintenance facilities develop robotic tools to assist aircraft inspectors Benefits Improved Detection Improved Repeatability Reduced Aircraft Downtime Electronically Retrievable Inspection Data Improved Safety for Inspectors
2. PROGRAM HIGHLIGHTS FINAL-PHASE DEVELOPMENT • MECHANICAL SYSTEM •
CONTROL
•
SENSORS
SYSTEM
• HUMAN-MACHINE INTERFACE
3. MECHANICAL SYSTEM DETAILS • MECHANICAL DESIGN OF FIRST-PHASE ROBOT
• SECOND-PHASE MECHANICAL IMPROVEMENTS • SYSTEM PERFORMANCE • LONG-TERM MECHANICAL DESIGN ISSUES
ROBOT MOTIONS
WALKING MOTION ALIGNMENT MOTION.
Spine assembly raised Spine assembly motion Spine assembly lowered Stabilizer bridge moved
5 CONTROL SYSTEM DETAILS • • • •
ON-BOARD ROBOT ELECTRONIC SYSTEMS GROUND-BASED ELECTRONIC SYSTEMS OPERATOR WORKSTATION COMPUTER SATELLITE EQUIPMENT ENCLOSURE
• control points (On=White, Off=Gray) • limit switches (On=Yellow, Off=Gray) •suction cups (Vacuum=Green, Ambient pressure=Red)
. CONTROL SYSTEM SOFTWARE
• NAVIGATION AND VISUALIZATION
6. SENSOR SYSTEM DETAILS. • EDDY-CURRENT SENSORS EDDY-CURRENT INSPECTION HARDWARE EDDY-CURRENT SENSOR CALIBRATION EDDY-CURRENT SOFTWARE
EDDY-CURRENT FLAW DETECTION •
VISUAL SENSORS AND VIDEO-BASED RIVET LOCATION ALGORITHMS. ROBOTICS INSTITUTE VIDEO RIVET LOCATION ALGORITHMS
CMRI VIDEO RIVET LOCATION ALGORITHMS NEURAL NETWORK RESULTS THIRD TRIAL: IMPROVED CAMERA SYSTEM RIVET LINE-FITTING ALGORITHM
•FEEDBACK SWITCHES
CONCLUSION The robot was able to achieve the following goals: adhering to and walking over an aircraft fuselage regardless of the orientations of those surfaces, acquiring eddy-current inspection data that appeared identical to manually deployed eddy-current sensors, and being remotely operated using video and sensor feedback provided to the inspector. Video camerabased automatic alignment and navigation were demonstrated in an auxiliary experiment using a separate robot moving on an aircraft skin surface.