Projects

SMITZH:
Automated inspection

Verification of the inner quality of a product, to check for defects, is a critical step in the aircraft manufacturing process. Automation of these safety insurance inspection processes is a major leap. TU Delft, TNO and SAM|XL have combined their expertise of robotics, sensors, data processing and aircraft structures to develop automated inspection processes. In the SMITZH project the first step of an autonomous NDT process was developed including autonomous part localisation, path generation and NDT execution.

Duration

June 2020 - June 2021

Partners

Purpose

In this project we concentrated on the automatic inspection of objects for small series, single pieces and critical materials. Consider, for example, the double-curved panels for the aviation industry. These panels are scanned using 3D vision, after which Zero Programming automatically generates robot paths. Pre-calibration of the robot relative to the panels is not necessary. This technique enables positioning objects arbitrarily, is contactless and the use of water to perform non destructive testing is avoided.

Automated inspection

Through its 3D camera, the robot can detect its position and orientation relative to the object of interest throughout the inspection process.

Eliminating the requirement of exact positioning of the part, shortening the cycle- and change-over time and making the system more robust against disturbances.
Using a MEMS Sensor Array, objects and products can be inspected for any damage without damaging or even touching them.
In contrast to classical Non Destructive Testing (NDT), objects can be inspected with the MEMS Sensor Array and without the use of water.
This inspection method attains an equivalent accuracy as conventional methods however with a increased speed of up to 6m2 per minute.

In the future, the goal will be to overlay the NDT scans over a CAD model, enable the robot to autonomously execute a second pass scan of suspect areas, and the incorporation of algorithms to rapidly determine whether a components passes the quality requirements, needs repair or should be scrapped.

Take a look at our showcase video's under Media to find out more.