Deep Learning, 3D, Lenses on Display at Automate 2022

Automate 2022 was live and in-person in early June, following on the heels of a pandemic-induced virtual event in 2021.

The Association for Advancing Automation, or A3, (Ann Arbor, MI; www.automate.org) organized the show, which was held at Huntington Place, a convention center in downtown Detroit.  

There were no requirements for proof of testing and/or vaccination—a sign that the industry is emerging from the COVID-19 pandemic. Many people were there without masks.

There were more than 500 vendors with technology on display and attendees ready to hear about their products. For the machine vision industry specifically, some conversations centered around the role of deep learning, where the goal is to automate difficult tasks—such as detecting colors or classifying objects—with minimal time spent training algorithms.

Product developers also talked about 3D imaging, logistics applications, and incorporating vision solutions into robotic processes.

Cognex

Cognex (Natick, MA, USA; www.cognex.com) was one vendor at Automate with deep learning tools. It featured its ViDi EL Classify, which includes pretrained algorithms that can learn new tasks, such as classifying gears or identifying parts with defects, based on five-to-10 learning images per class.  

“What used to take four months of programming, now takes minutes,” said Eric Hershberger, application engineer at Cognex.

The deep learning tool is included in the In-Sight 2800 series vision system, which was on display at the booth.

SICK

At the Sick (Waldkirch, Germany; www.sick.com) booth, one display touted the value of 3D imaging in machine vision applications. It showed how the company’s 3D smart cameras can be used in applications to measure the volume of boxes—necessary information for space planning on pallets or trucks. “3D is a lot more accurate. 2D has a lot of limitations,” Jeff Koncelik, sales marketing manager at SICK, explained at the booth.

Sick’s 3D cameras include the Visionary-T and Visionary-T AP (programmable) series, which use Time of Flight (ToF) technology, and Visionary-S series, based on stereo vision.

VS Technology Corporation

VS Technology Corporation (Tokyo, Japan; https://vst.co.jp/en/), which is known in the semiconductor industry, wants to expand into food and beverage and logistics. Representatives from the company were touting lenses, including the VS-HVA series of fixed focal-length C-mount lenses, which are compact and lightweight, ranging in weight from 126 g to 142 g.

The lenses are ideal for robotics, according to Shusaku Aimoto, NBD sales manager and application engineer. For applications involving a robotic arm, he said, a “challenge is getting a camera that isn’t too big or weighs too much.” That’s because engineers and integrators don’t want the camera solution to impede the robot’s movements.

These are just a few of the companies featured at the show, which was educational and enjoyable. It was refreshing to meet with people in person instead of via a computer screen.