Now factory robots may see better than you with 3D vision

Industrial robots, or factory robots as they are more popularly known, are all set to take over more responsibilities in the coming days. This was a natural progression in technological evolution, expected to happen in a few years. But COVID-19 has accelerated the process, increasing demand for robots to take over work from humans so that businesses can continue operations without health concerns.

Recent developments in industrial robotics have focused on computer vision that enables robots to perceive what's in front of them. This is an integral part of industrial robots for several applications. Developments in artificial intelligence have helped robots to improve vision, become more efficient than ever before. Here's an overview of some of the recent developments from major robotic manufacturers.

The role of 3D vision

The 3D vision enabled automation of pick and place manufacturing processes by industrial robots is one of the most notable developments. The

process is called 3D bin picking, where parts are fed in a tote bin, and the robot uses a 3D vision to pick and sort. 

"Picking of randomly placed parts in a bin across assembly, inspection and transportation phases has a strong reliance on people's experience and senses amid deepening labor shortage, the new normal safety guidelines of social distancing and increasing labor costs of recent years," explains Swaminathan Ramamurthy, GM of OMRON Automation Centre& Robotics at OMRON Asia Pacific. "The 3D bin picking automates the picking of parts by equipping the robots with "eyes," that have high-speed 3D vision sensing capable of instantly recognizing the shape, position as well as the orientation of parts." 

Ramamurthy added that OMRON has recently developed this 3D vision sensing technology, which, coupled with our 6-axis robots, enables high-speed, high-accuracy three-dimensional object recognition and provides a compact and lightweight vision sensor that can be mounted onto six-axis robot hands.

Increased efficiency

The robotic manufacturing giant ABB offers a range of easy-to-use vision systems for robot guidance and inspection tasks. Machine vision hardware and software tools are designed to reduce engineering time, increase flexibility, and enable faster and more efficient production, picking, and packing.

"Our vision system guides ABB robots for material handling tasks, identifying workpiece locations, and guiding the robot with high precision in the work cell," said Subrata Karmakar, President of Robotics and Discrete Automation at ABB India. "The software is our market-leading robotic software for vision-guided random flow picking and packing applications. It features a powerful color vision system that can support up to 10 cameras for accurate position guidance and feature inspection, enabling faster and more efficient picking and packing."

Karmakar added that the company's new robotic 3D vision and metrology system scan defects in manufactured products. The 3D Quality Inspection system can detect faults that are less than half the width of a human hair to enable quick and accurate quality checks. Enabling a higher level of quality control will help customers reduce expensive reworks and waste.

The role of AI and ROI

In a nutshell, robots are using sensor technology to perceive workspace. Sensor technology allows robots to understand the environment. With AI, robots can sense more in their environment and add more value to the work.

“So the robot that moves around uses AI, it uses computer vision, it has stereoscopic cameras, similar to what you would see in a Microsoft Kinect, gets a 3D view of the space around the robot," explained Stefan Nusser, Chief Product Officer at Fetch Robotics. "With that, it opens up this opportunity to understand the workspace and create value. For example, it can detect when something flops or detect when a forklift is in an aisle where it's not supposed to be. These are all these observations that the mobile robot makes as it moves around. That is, that is where the value is to come out with the AI.”

Conclusion

Much of the recent developments in industrial robots revolve around how machines can better see and understand objects around them. This is achieved through a combination of cameras, sensors, and software that analyzes data captured. The software is where AI comes into play, and it is in this area, we may see more developments in the coming days.