DENSO Tackles Cobots, Large Robotics, and “Entry” SCARA Robots at ATX 2020

February 13, 2020 by Kate Smith

DENSO showcased three new big strides at ATX 2020.

This year at ATX 2020, DENSO showcased how they're stepping into three new spaces—cobots, large robots, and more affordable robots.

Control Automation got to speak with David Robers, DENSO Robotics Sales Manager and manufacturing engineer, about the evolutions the company is pursuing and the trends they represent.

Robers believes that DENSO has a distinct advantage in that the company made their own robotics for their own automotive manufacturing use before branching out to provide them for others. When it comes to DENSO's relationship with their own robots, he says, "Our message is that we use them. [When we release a robot], we've been running it in our plants, testing it out, getting cycles on it and making sure that it meets that reliability. So not only our internal R&D testing, but actually they're putting it in our plants and using it in production."


Cobots and the Global Drive for Automation (AKA Why Japan Has Been Slow to Adopt Cobots)

One of the major changes for DENSO showcased at ATX 2020 was their new cobot, Cobota.

DENSO joined the "craze" of cobots a little more slowly than some. The reason for that, Robers says, is DENSO's roots in Japanese automation.

"Japan, in general, is slow to get into cobots because Japan's been automated for 50 years. The US is really behind around the globe. We're about 30% automated in manufacturing. Europe's about 60-70% and Japan's like 80-90% percent automated."

This means that the Japanese are working on second-generation automation, which can decrease the immediate appeal of cobotics, which are (so far) limited in size, speed, and load capacity. "Something that moves slowly and doesn't have a big payload and isn't that accurate? Really, they don't even understand it."

The reticence of Japanese robotics makers to embrace cobots, of course, raises the question of why cobots are so sought after to begin with. Robers explains that the key is integration, from grippers to vision systems. Cobots sidestep the need to invest in the development of safety wires, grippers, etc.



Gesturing to the Cobota demo at the DENSO booth, which has been diligently picking up and dropping pens from a bin, Robers tells us that no one taught the device what a pen was. No one uploaded a CAD model. It was exposed to samples of pens and can now be placed anywhere to pick up pens. But this functionality would theoretically be possible in industrial robots, which would also be more customizable. Its edge is its accessibility. 

It has what Robers refers to as easy "drag-and-drop" programming that's unusually simple for a cobotic pick-and-place or vision finding platform. This is also programmable via the DENSO industrial programming language, WINCAPS.

Additionally, "It's got a gripper on it. You just have to make sure that you put some fingers on it. The only thing that really is custom at that point is the interface between your part, which could be anything, and the robot end-effector, itself. So they just 3D printed fingers in this example. Now you start programming—plug it in and start programming. That's not how industrial robots work... [With cobots], you could be up and running for a simple application very quickly.

In this way, a situation that used to require a minimum of $80,000 now may only cost $20,000—a point Robers says is close to the maximum amount most people are willing to risk in trying something new. Since the Cobota is available for $14,000 without a gripper and vision to $20,000 with the gripper and vision, Robers argues that this makes cobots even easier to integrate into a system with minimized risk.

"This will be the gateway drug for automation."


Large Robotics 

Another new thing for DENSO on display is big robots. Robers characterizes DENSO as experts in small to mid-sized assembly robots, specifically of the SCARA and six-axis varieties, though various customers have apparently asked about the large robot option for years. In the past, DENSO would need to point to a competitor to finish out a production line for palletizing and large loads—but that's no longer the case. 

Scheduled to release in July of 2020, DENSO's new VM • VL series of robot was on display—sporting a photo of Al Pacino's Scarface on a sign that read "Say hello to my BIG friend!"



DENSO's robots are generally sleek in design, an aesthetic that also serves functionally. 

"If you look at most of the larger robots," says Robers, "they have external motors and external wiring harnesses and things like that. But they get dirty. They get snagged on other devices and accessories." He goes on to explain that these features, such as a valve stack—usually housed in a valve pack mounted on the outside of the robot—are contained within the DENSO robot's arm.

There was a learning curve for DENSO in designing large robots. In the past, two (or even one) person could move a DENSO robot. Members of this new line weigh about 1,400 pounds, so there are obvious logistical differences.


Integrating Automation: An "Entry" SCARA Robot 

The third new space DENSO is moving into is low-cost, "entry-level" SCARA robots, the DENSO LPH series. Going back to the concept of customers with a threshold of tolerance for risk, this series has been posed as that "gateway drug" for those looking to experiment with lightweight applications without investing a huge amount. 



Some may ask why anyone would reach for the more expensive SCARAs if this one is available. "This is a four-to-five year duty cycle," says Robers. "The robots we're used to making are 12 to 15 years."

There are situations, Robers says, where manufacturers may choose to emphasize reduced cost over longevity more strongly. He cites the example of cell phone manufacturers, who may be less concerned with longevity because of the shorter design cycles for their products. Some of these customers may "tear their whole line out every 18 or 24 months" and build a new one when the newest version of their product demands it.

It's also possible to shift to the larger SCARA robots after the "entry" period has passed. Besides the re-deployability of DENSO robots, Robers asserts that the same program that the entry-level SCARA runs on and put it on the bigger model using the same software structure, just using a different structure.

Another new relevant initiative is the ability to control some of DENSO's standard SCARA robots off of a single controller. Robers breaks down the cost-savings as around $12,000 when robots are deployed in these pairs.


One for the Controls Engineers

Robers wraps up our conversation with something he believes controls engineers would be pretty excited about: a completely open architecture for programming. Aside from PacScript through WINCAPS, it's also possible to program in LabVIEW, as well as utilizing instruction sets from Rockwell, Omron, Siemens, Beckhoff, etc. 

"You could program directly out of your PLC software if you want. There's a way to program our robot directly out of just about any programming language," he says, rattling off Python, C++, Visual Basic., Python—"whatever their favorite programming language is to control the robot. And generally, those are all free libraries."