This article was first published on October 2, 2020.
We’re pleased to announce that we have received a $250K grant from AMGC (the Australian Manufacturing Growth Centre) to lead the development of an AR (Augmented Reality) based Quality Control and Inspection system, initially for use in Printed Circuit Board manufacturing.
The solution will reduce equipment and engineering costs and increase operational efficiencies for PCB manufacturers and assemblers of electronics. An inspection system that uses AR likes this does not currently exist.
“Ultimately we’re making a system that will be the equivalent of a Google search for PCB hardware design. All stakeholders across design and manufacture can gain insight into the what, where, when, and why of a design through an AR based user interface.” – Matthew Adams (CEO, Tekt Industries)
How it works
The core of the system is based around a new paradigm for display of AI inspection feedback and design information delivered as a result of deep machine learned board identification. This information is rendered to an AR headset while being overlaid on the physical PCB in front of the operator… magically giving the user X-ray vision and… other super-powers.
The solution includes a camera and vision processing module that uses AI (Artificial Intelligence) to recognise a PCB’s features. This inspection system has access to the design files associated with the board, as they are stored in the Cloud Asset Management Portal.
The inspection hardware compares this downloaded ‘digital twin’ (the design file) with the physical PCB in front of it, identifying design features and quality control issues. The operator can then view this information via MR (Mixed Reality) smart glasses, in this case: Microsoft’s recently released Hololens 2 product.
Key benefits
The modular system can be placed anywhere in the production line or it can be used as an offline production evaluation system. This will allow for flexible deployment.
It will be an Industry 4.0 connected system, designed to quickly access design files associated with the board for comparison from a central cloud repository. This will eliminate the need for an operator to load necessary information into individual machines. The machine itself downloads the necessary assets.
It’s estimated to cost between $25K and $50K, making it significantly cheaper than existing solutions and more affordable for SMEs and large enterprises seeking to improve their quality control and inspection capabilities.
The Augmented Reality component of the system will allow an untrained operator to perform the work of an expert through AR visual cues and live heads-up information.
System components
1. The Augmented Reality Headset
2. The Cobot Robotic System with Custom Vision End Effector
3. A Conveyor Sub-system for PCB Transport
4. An Environmental Monitoring System
5. AI Accelerated Central Processing Hardware
6. Cloud Asset Management System
The take home
Working with our project partners, we’ll develop a solution that increases efficiencies in production and engineering for local and international markets.
We are grateful to AMGC for their network access and financial support, as we deliver this exciting new solution for the manufacturing industry.
Project partners: The University of Melbourne, Entech Electronics, Microsoft, Mobile Automation
