A machine vision and imaging conference in September offered guidance on artificial intelligence-based machine vision perception, vision-guided robots and visual language models, among other topics. The Association for Advancing Automation also offers machine vision certifications.
2025 A3 machine vision conference insights
- Machine vision experts from Intrinsic, Fanuc America, Osaro, MVTec USA and Apera AI were among those sharing knowledge at the September A3 event, “Focus: Intelligent Vision & Industrial AI Conference.”
- Trends from the conference include artificial intelligence-based machine vision perception, vision-guided robots and visual language models.
- Association for Advancing Automation is helping those interested in vision and imaging by offering machine vision certifications.
The integration of machine vision and artificial intelligence technologies are transforming how industries operate, said the Association for Advancing Automation (A3), about its A3 September machine vision conference in Seattle, “Focus: Intelligent Vision & Industrial AI Conference.”
Rapidly advancing technologies boosting machine vision performance was discussed and demonstrated at the conference, helping attendees increase productivity and quality, A3 said.
Trends in machine vision, machine learning
In a session called “Foundation Models and AI Capabilities in Robotics Software Platforms,” Torsten Kroeger, chief science officer, Intrinsic, said autonomous industrial robot systems can be programmed without deep expertise in AI, computer vision and programming. Software platforms can abstract robot and sensor hardware assets. Foundation models for robot vision can be prompted with computer-aided design (CAD) and image data once (so called “one-shot learning”) to create robust sub-millimeter pose estimation capabilities for a wide range of use cases. Robotics and automation solution builders can create value for customers by deploying robots that do not need to be programmed, Kroeger said.
Advances in vision-guided robotics
Another session at the conference covered advances in vision-guided robotics. AI is helping vision-guided robotic material handling, assembly, machine tending and other applications as the following speakers explained:
- David Bruce, engineering manager, Fanuc America
- Gemma Ross, vice president of operations, Osaro
- Heiko Eisele, president, MVTec USA
- Sina Afrooze, CEO and co-founder Apera AI.
Bruce covered “Practical AI Uses for Vision Guided Robotics,” cautioning about hype surrounding modern AI techniques. It is possible to use AI techniques to augment what is possible with industrial and power and force limited robots. Technologies that can be applied include ROS 2, re-enforcement learning and AI for vision segmentation to assist robots with segmentation tasks.
Ross’ presentation was “From Research to Reality: Commercializing AI Robotic Picking at Zenni Optical.” Ross pointed out that commercializing advanced AI-driven robotic picking involves overcoming real-world challenges often absent in controlled lab environments. At Zenni Optical, AI-driven robotic grasping and vision technology moved from the research lab into a high-throughput commercial fulfillment environment. In the implementation, challenges included barcode visibility, reflective surface scanning, precise object handling, integration with multiple automation components and real-time machine learning adaptation. Measurable improvements in operational performance resulted, Ross said.
Eisele covered “Scalable Bin Picking with 2D Cameras: A Hybrid Approach Using Deep Learning and 3D Vision.” Automated bin picking often fails when parts are shiny, overlapping or geometrically complex. Using a scalable vision technology integrates deep learning and classic 3D algorithms. Single 2D area scan cameras can reliably detect and localize objects in 3D, eliminating the need for expensive hardware setups. Synthetic training data can replace manual annotation to streamline rollout across product variants. System architecture should focus on practical considerations for real-world automation environments, Eisele said.
Afrooze looked at “Deploying AI to Retrofit Manual and Vision-Guided Robotic Processes to Unlock Factory Performance,” with case studies from the industrial manufacturing industry involving the deployment of artificial intelligence (AI) for vision-guided robotic applications.
The projects were retrofits of manual processes performed by operators, robotic processes without vision or cells using conventional vision technologies and demonstrated the following trends:
- The main driver of process retrofitting is quality and eliminating stops to production.
- Vision systems for robotic guidance require specific process and mechanical redesigns.
- Work cell retrofits can deliver expected performance increases.
- Project teams may need new skills to apply and use vision-guided robotics.
Association for Advancing Automation vision certifications
In related news, A3 offers machine vision education and certifications. The A3 Certified Vision Professional program covers machine vision and imaging concepts and techniques and is offered in two levels of certification: CVP-Basic Certification and CVP-Advanced Certification.
A3 basic machine vision and imaging courses are:
- The Fundamentals of Machine Vision
- Beginning Optics for Machine Vision
- Beginning Lighting for Machine Vision
- The Fundamentals of Camera and Image Sensor Technology
- Image Processing Fundamentals
- Vision System Design
A3 advanced machine vision and imaging courses are:
- Advanced Optics for Vision
- Advanced Vision Lighting
- Advanced Camera and Image Sensor Technology
- 3D Vision System Development
- Introduction to Machine Learning
- Designing Linescan Vision Systems
- High-Speed, Real-Time Machine Vision
- Advanced Vision Guided Robotics
- Non-Visible Imaging: Infrared Technology and Applications
- Advanced Color Machine Vision & Applications
- Metrology and 2D Calibration Techniques
- Advanced Image Processing
- Advanced Vision System Integration
Find more information about A3 machine vision and imaging certification at https://www.automate.org/vision.
Edited by Mark T. Hoske, editor-in-chief, Control Engineering, WTWH Media, from an A3 press release, [email protected].
Keywords
Machine vision trends, A3 machine vision certification
Consider This
What techniques are helping your machine-vision implementations?
You also might like
Control Engineering provides more machine vision knowledge.
https://www.controleng.com/mechatronics/vision-and-discrete-sensors
