Machine Vision Makes Collaborative Robots Even More Flexible

Robots are finding their way into every niche of the manufacturing industry because they can save humans from the 3Ds — or work that is dirty, dull, and dangerous. But at their essence, robots are automated devices made for material handling — essentially metal hands designed to move an object from one location to the next. But what good are hands without eyes and a mind to guide them?

For years, machine vision systems have provided the technology necessary for robots to see. These vision guided robot (VGR) applications have guided traditional robots for pick-and-place, machine tending, assembly, and even complex bin-picking operations. Today, this same technology is helping cobots — collaborative robots that can work alongside humans without safety fencing and other measures — to grow from only a few percentage points of the robot market to an anticipated 35% of all new installations by 2025.

COVID-19 Increases Need for Flexibility

Robots without “eyes” can be incredibly useful at solving manufacturing jobs. But consider all the steps required to help a robot do a simple task without the use of machine vision guidance.

The main challenge is that the object has to be presented to the robot in the same orientation and location every cycle. This means installing costly custom fixtures, jigs, and hardware such as bowl feeders so that randomness is eliminated, and the robot can predictably and repeatably pick the object at a given location.

Now, what if the production line has regular turnover, with each product requiring new fixtures and tooling? What if it’s a mixed production quality assurance or packaging line? What if the work cell needs to be capable of handling several types of objects arriving in different orientations?

In these scenarios, with flexibility and cost in mind, designers should consider the benefits of adding robot vision — i.e., a camera and processing unit with associated machine vision software — tor production lines. Vision eliminates the labor and capital costs associated with the addition of new hardware and fittings, adding flexibility to operations and enabling the line to quickly switch between different products. And vision dramatically reduces the time required to move between one product and another — a massive boon for busy manufacturers, especially in today's demanding, customized-production environment.

“Quality Magazine's 19th Annual State of the Profession Survey conducted in March of this year found that two in five companies are using or planning to use robots, and half of those are already training their existing workforce on robot technologies,” says Joe Campbell, Strategic Marketing and Senior Manager of Applications Development at cobot specialists Universal Robots. “These pre-existing trends have been given extra impetus by the global COVID-19 pandemic. The manufacturing sector is adjusting to new operating conditions, decreased labor availability and increased economic uncertainty. As a result, the demand for flexible automation platforms has intensified.

“While traditional automation systems are ingenious,” Campbell continues, “they can also be expensive and rigid. By contrast, cobots can be assigned to one set of quality inspection tasks in the morning and reassigned to a completely different set of quality inspection tasks in the afternoon. One vision system and be swapped out for another in minutes. And so, traditional 'one-size-fits-all' approaches to automation are increasingly losing ground to customizable cobot platforms.”

To give users the flexibility to choose vision systems that are best for their application, Universal Robot has an extensive ecosystem of partners that offer vision systems tailored for specific purposes, such as pick & place and machine tending to smart 3D systems for vision guidance and quality inspection.

What Collaborative Robot Camera is Best?

Robot cameras come in three main flavors: 2D, 3D and 2.5D, according to Kristian Hulgard, General Manager of OnRobot’s America’s division. 2D cameras are the cheapest of the bunch, but are the least versatile. “Typically, 2D cameras determine length and width (X and Y axis), but are unable to determine height, which limits the number of applications they can support. On the plus side, they are reliable within these constraints,” she said. “3D cameras provide your robot with all the visual information it could possibly need, across all three axes and incorporating object rotation. This functionality comes at a price however, since 3D cameras are the most expensive cameras out there and they also have a reputation for being difficult to integrate and operate.”

Hulgard adds that 3D cameras can bring complexity and reliability issues that make many manufacturers reluctant to embrace the technology, despite its powerhouse capabilities. Instead, she points out that 2.5D cameras occupy a sweet spot between 2D and 3D cameras, both in terms of cost and capabilities. Capable of determining the height of objects, 2.5D cameras are ideal for scenarios in which objects differ in height and when items need to be stacked.

Vision Increases Collaborative Robot Density, In-Line Inspections

Other cobot suppliers, such as FANUC America, prefer to offer customers a fully integrated vision solution sourced from the robot supplier, such as FANUC’s iRVision, which operates on both their cobot and traditional robot product lines.

“What separates traditional robot vision from collaborative robot vision? Nothing really,” explains David Bruce, Engineering Manager, General Industry & Automation Segment for FANUC America. “Since cobots are industrial robots with intrinsic safety features built in as part of their design, the principles of using machine vision for guiding a regular industrial robot or a cobot are the same. That being said it is likely that cobots will be used more for robot-mounted MV inspection. Cobots are well suited for this because it often involves an operator loading a part into a check fixture, and if this fixture is going to be inspected by a cobot then the entire process becomes easier [because humans and cobots can work together].”

Cobot inspection stations can be either in-line, inspecting every part moving along a relatively low-volume production line, or they can inspect every tenth, one hundredth, or one thousandth sample from production line at an off-line, operator-fed inspection station, says Bruce. “In this type of setup, an operator can load the part to be inspected without having to interface with the robots’ control system or enter through any kind of fencing, which is a very desirable attribute for these kinds of systems. Sometimes guidance is incorporated with these types of systems if the part is not placed in a fixture. The cobot will first use a robot-mounted or fixed-mounted camera to locate the position of the part and then use this information to offset its inspection path to ensure a proper and accurate inspection is performed.”

Machine vision systems also have a place off the robot arm, according to Bruce. By installing fixed-mounted cameras and software, similar to VEO FreeMove, vision systems can monitor robot and worker position, allowing more cobots to be colocated in a small space without concerns about colliding into one another or human workers.

The combination of machine vision with industrial robots, often referred to as VGR, either collaborative or regular, has been on a steady rise for several years. “All the major industrial robot companies offer some form of easy machine vision integration,” says Bruce. “Whether for machine guidance or part inspection, the combination of machine vision with industrial robots is really one of the necessary technologies to enable the automation of many of the tasks that [otherwise] are simply not possible to automate.”

At Comprehensive Logistics in Youngstown, Ohio, a UR10 cobot equipped with a Keyence CA-HX200 C Lumitrax vision system is deployed on the fast-paced engine assembly line, checking the correct assembly of the plug-ins for the steering gear wire harness. The task is mission critical. If there is an intermittent connection caused by a connector not fully seated and locked into place, it can cause a loss of the power steering, which is a level 8 on the severity scale - meaning a life-threatening failure mode. Courtesy of Universal Robots

Flex-N-Gate, a supplier of bumpers, exterior trim, lighting, chassis assemblies and other automotive products, uses inspection systems at their Ventra Ionia, Michigan plant to ensure product quality. To help improve these processes, reduce costs and save floor space, Flex-N-Gate turned to FANUC for a collaborative robot solution with vision, leveraging FANUC America's 24/7/365 service network to support their cobot systems for a completely successful integration.  Courtesy of FANUC America.

The newly released OnRobot Eyes is a 2.5D camera and vision system that can be used on all major collaborative and light industrial robot arms. Designed with ease-of-integration and -use in mind, Eyes brings all the benefits of robot vision to manufacturer operations along with some unique capabilities and features. Courtesy of OnRobot.