Enabling Machinery Intelligence with 3D Vision
Here at Facteon, we’re on the leading edge of automation, robotics and software systems. We’re also working to explore the potential of 3D vision systems. Our research has found that whilst 2D vision is of great value, it doesn’t offer the technical edge of 3D vision. 3D machine vision allows production robots to better visualise their environment. It allows the machines to adapt to their environment, increasing flexibility, speed and utility.
Due to this strategic advantage, 3D vision is an integral part of Facteon's in-house research and development programme. Our R&D centre of excellence includes a team of specialists researching how 3D vision technology can improve our products and services as well as our customers’ manufacturing performance. In the future, we seek to integrate 3D vision systems with our machinery to provide greater flexibility and improved efficiency of operations. Facteon is also committed to advancing the technology while developing the collective pool of industry knowledge in the field.
Our R&D team is working to understand the science, engineering and mathematics that the technology is based on. Facteon prides itself on providing customised manufacturing solutions native to the customer’s factory environment. Our future success in this area hinges on our ability to comprehend the science behind the technology, as this will allow us to better identify how 3D vision systems can potentially enhance the operations of our customers. To develop this knowledge, our in-house experts work closely with local university specialists.
Facteon is exploring the potential of integrating 3D machine vision in processing systems such as aligning, locating and dimension inspecting. We’re currently exploring the potential applications of defect detection, texture and material classification, assembly verification and deformed part location and character reading. This allows for the generation of as much product information as possible, at production line speed, with limited human input.
Facteon is motivated to understand this developing technology as traditional techniques such as displacement probes are limited in scope. This is what makes 3D vision such a valuable tool. Facteon customers are demanding precise data without a costly production line slow down as the measurement is taken.
To build sophisticated industrial 3D vision systems you need the right hardware, but you also need the right software. Our R&D team is working to better understand the software side of 3D vision. There are many software options available such as Cognex’s VIDI for image classifying and quality inspection. Halcon for pre-trained OCR functionality. Then there’s Microsoft’s CNTK for deep learning applications. However, our research has found that value lies in the use of these software tools, combined with artificial intelligence, deep learning, neural networks and proprietary IP and algorithms.
The software is supported by the highest quality cameras and structure lights. This may involve moving stereo cameras or revolving the product under investigation. Structured light allows for enhanced analysis of relatively flat, non-descript surfaces.
Above all, we seek to make our production machinery more intelligent. 3D vision systems show great potential to enhance the future machinery and software systems in factories. While we strive to continue bringing industry-shaping solutions to market, we also seek to contribute to the collective pool of industry knowledge in the area of 3D vision systems. It’s also important to note that Facteon’s manufacturing roots shape our approach to researching this area. Rather than taking a largely theoretical viewpoint, we seek to understand how technologies such as 3D vision will be shaped by overarching industry trends, including Industry 4.0 and Industrial IoT.
I’ll be back with another blog post in the coming months to give an update on how our 3D vision systems research is progressing. Watch this space.