Automated Robotic Inspection
Event address
Learn how to make a robot perform quality inspection tasks.
- Hardware: viewpoint determination, path planning, collision avoidance
- Software: image acquisition, annotation and building an AI-model
In the course, participants will be tasked to perform a robotic inspection of a 3D printed dragon. To this end, we will create a robotic inspection path based on the CAD-file of the object and our in-house viewpoint optimization tool. The participants will then upload the program to the robot and perform the inspection. The obtained images will then be used to train an 'AI' feature-detection algorithm. To this end, the participants will annotate the images and perform the Ai-training. To conclude the session, the participants can implement their model on the robot software and check out its performance in a live setting.
Participants are allowed to bring their own object for inspection, as long as they send us a CAD-file of it, a week in advance.
Your Teacher
Sam Dehaeck is a senior researcher at Flanders Make specializing in computer vision and optical setup design. He has worked more than 10 years in the design of space experiments applying optical interferometry on evaporating fluids. The last 5 years, he has been focussing on deep-learning based machine learning techniques and its application in production environments. Topics researched in this respect deal with operator recognition for ergonomic ends and viewpoint optimization for robotic inspection.
Practical Information
Friday February 21 2025, Flanders Make Kortrijk
8:30 - Welcome Coffee
9:00 - 17:00 - Workshop day, lunch and coffee breaks included
Language: English
Participation is free of charge, please register via the form below.
Registration
This training is part of the Digital Europe project "RebootSkills".
