ABB & ABICOR BINZEL Discuss Optical Seam Tracking

ABB & ABICOR BINZEL Discuss Optical Seam Tracking


– What’s going on, guys? This is Matt Sciannella
dealing with Matt Conklin, welding applications
engineer for ABB Robotics, and Scott Huber, key accounts manager
for sensors and robotics at ABB and Abicor
Binzel respectively. We’re here at ABB Robotics
in Auburn Hills, Michigan, talking seam tracking for
arc welding applications. Scott, I’m gonna
to start with you. First question I have
about seam tracking. Tell me how seam
tracking has grown from its first inception
into today’s seam trackers. – 20 years ago, the CPUs weren’t as
strong as they were, they couldn’t handle much data. Today, we’re able
to take that data, process it faster, allowing for closer lookaheads, shinier materials, different weld joints. The memories in
these are becoming, so it can hold more jobs, look at different applications. We’re allowed a
three-line technology. So we’re able to get
that data back in, process it in milliseconds, be able to get the
data to the robot and make those corrections
and offsets as needed. – What should people
know about seam tracking if they’re looking
at it right now for their robotic
welding application? – With optical seam tracking, it’s there to enhance
what you currently have or what you would like to do. It’s not the fix-all cure, but it can do a lot. It can definitely help in parts that are not
fitting up correctly, that they vary a little bit. But they have to
vary within reason. We can’t just have parts that are scattered
all over the place, that we just say “Hey, we’re gonna set
them down and weld it.” There are some limitations, but overall the process is
very, very good and reliable for enhancing what
you currently have or what you would like to do. – There’s a lot of seam
tracking options out there. Not all of them
are made the same. Some of them vary in
features and spec. So Scott, I’ll start with you. What are some of the
evaluation criteria you tell a person who’s
looking at seam tracking? Basically, what separates
optical seam tracking solutions from one another? – So when you’re looking at, that’s a great question too. ‘Cause when you’re looking at going to optical seam tracking, a lot of people will
look at their parts. They’ll have off-location welds, they’ll look at tooling
that’s expensive and “Can I set back on
my tooling a little bit?” So when evaluating going
to optical seam tracking, what you have to look at is what does my weld
joint look like, what’s my material type, what is my end goal. ‘Cause like Matt alluded to, it doesn’t fix your
non-weldable product. So if your part’s not
weldable to begin with, if you got large gaps, if the material fit up
isn’t allowed to be welded, optical seam tracking
is not gonna fix that. What it’s gonna do
if you have a part or the weld join is just
moving in different locations, ’cause today’s robots
are highly repeatable. They go to the same
spot every time. What optical seam tracking does is allow the robot to
have a set of eyes per se. So it’s gonna look at the joint, it’s gonna say “Hey, this
is where the joint is, “this is where I was taught,” and then it’s gonna
adjust to that point. So you really have to look at does optical seam tracking
apply to that process. I’m gonna let Matt take over, talk a little bit about
the robot side of it. But that’s what we’d look at to even say yes you can
put an optic on this. Then we go to our robot partners and say “Hey, this is
what we’re looking at. “It can do it. “What do you guys need?” – So a lot from our end, we start looking at, okay, is there access? Can we fit the torch
and the camera in there? Do we have to use
a different torch? Binzel has has a
great line of torches. And do we have to
mount the laser in a different orientation? One way that we like to qualify these
different leads and things is through one of our
tools, RobotStudio, which is a offline
programming tool. We actually have 3D
CAD models from Binzel. We actually bring them in, snap them to the robot, and start checking reach offline even before the
tooling is built. And where we kind of start
looking at differences here is, through the arc seam tracking
is one option that we can do but it has its limitations also just like every other process. Thinner materials
doesn’t do as well, aluminum doesn’t do as well, and that’s where the TH6 and a lot of the other
models really excel. On the shiny materials. On aluminum, does a great job. So there’s a lot of
opportunity there for optical seam tracking. – Why is optical seam tracking, in your opinion, the most effective
solution for seam tracking? – Well, the aspects of
optical seam tracking, it’s zero contact. You don’t have to
touch the part. We’re able to have a
stand off distance. We’re able to get
real-time data. We’re able to do the
different material types. Aluminum is one of the
hardest things to do through the arc seam tracking. Thin materials. So optical seam tracking
gives us those ability to track the materials that we were never be
able to track before. Aluminum diamond
plate for instance. We’re able to track that in
a butt joint and a lap joint. Historically, that was
not a tracking solution. But today’s
technology and the TH6 allows us to be able to do that. The zero gap scenarios. So if you got zero
gap butt joint, TH6i is gonna be the solution
for you at that point. We just want to look
at what material it is. Optical has given
us a lot of advances versus touch sensing. Touch sensing is slow. Optical seam tracking, we can typically
find a weld joint within about a quarter second. And with touch sensing, you’re be about three to five
seconds per search vector. So if you go to put a 2D, you’re under that 10-second. If you’re doing a 3D touch up, the 15 seconds just
to find the start. Then typically you’d find that. If it’s a straight line, you’re gonna find the start, you’re gonna find the end, connect the dots. If you’re going to
through the arc, you’re gonna have to
have a weave on there which is very slow. Optical seam tracking, what happens, we find the joint, we’re able to track it. Typically, you’re able to
increase your travel speeds. So better production rates. You’re able to stay
in the weld joint. You can always usually
jump up a wire diameter because now I’m not as finicky. I can go faster with everything. So optical seam tracking
has a lot of benefits that I would say
that non-traditional or non-typical anymore. We find a lot of
situations where through the arc
was the solution. And optical seam tracking is kind of trumping
that going forward. The systems are
becoming smaller. We’re able to get in. And with technology right now, we’re able to
pre-scan some parts, drop off the scanner, come back in, and run those travel
speeds at higher rates. – We talked a little bit about comparing seam tracking
to other solutions. Tell me from a
communication standpoint, joints and materials standpoint, how seam tracking, optical
seam tracking in particular, how does that compare to
solutions like touch sensing or like through the
arc seam tracking from a performance standpoint? – Right, yeah. They all have their limitations. When you look at it as a whole, optical seam tracking has
a lot more opportunity. If you’re doing a lot
of different joints, thin materials, thick material, multi-pass, single pass, different materials in itself whether it be steel or aluminum, it really has the most variety
of things that it can handle. When you start
looking at WeldGuide and things like that through
the arc seam tracking, the limitations
stop at aluminum. We can’t WeldGuide
with aluminum. In certain stuff, we have a little bit
harder time with pulsing, things like that. So there’s limitations there. The same with touch sensing. When you start touching either the cone or
the wire to the part, if it’s a heavy mill scale, we run into issues there with not getting an
accurate touch sense. There could be
things on the nozzle, the wire could be bent. There’s a lot of factors
that go into play there. To get a really good,
clean touch sense, you need to first send it over
to the torch cleaning station and snip the wire. Which, as Scott has
alluded to, costs time. So there’s a bunch of
different things that add up that help us decide on what is the best
solution for the customer at the end of the day. – Where optical seam tracking
has some of its limitations, it has to be in field-of-view. So if it can’t be field-of-view, it has to be the
right torch angles, has to be the right
camera angles. Touch sensing, as long as you can reach the
thing and get a good hit, you can get the
read back for it. We’re looking around tooling, physical reach of the robot, can I physically get there? Those are the limitations
that you’re looking at. You got a chunk of steel sitting on the front
side of your torch or somewhere around the
peripheral of that torch, it just gets in
the way sometimes. – Wanna talk to you
guys about cycle time with optical seam tracking. For any manufacturer
out there time is money. You want to be
welding and on the arc as much as possible
with your robot, right? So tell me how optical
seam tracking does in terms of cycle time against other seam
tracking solutions. – What we find is touch sensing versus seam
finding with an optical laser. Touch sensing, like
we talked about, is very slow. It’s three to five
seconds per search vector. Optical’s about
a quarter second. So there right there, you can save anywhere from
10 to 15 seconds per weld. Sometimes if you’re doing
arc starts and arc ends, you’re up around 20
to 30 seconds just
for searching alone. With the optical seam tracking, you never have to
search your ends. You’re always
searching your starts. – If we can keep the part within the field-of-view
on the camera, we have a pre-process buffer. So we can actually buffer the distance from
the laser to the arc, and then we can strike the
arc at where we want to and be in position already. Another way that we can do it, if we’re not gonna be able
to find it within the window, we can actually do a
search and find the edge and then start from there. And then as Scott has said, that is all done much quicker
with optical tracking. – Is there different optical
seam tracking solutions? Are there ones that
are faster than others, maybe ’cause they do a
real-time joint scanning compared to pre-scanning
or anything of that nature? – Yep, so there’s other optical
solutions on the market. One thing that separates
TH6D is three lines. So we got three lines of
separate data coming back. We’re averaging those together, giving us a more reliable path. The information is
three times better than the other
solutions out there. So with that
processing coming in, we’re able to make sure that
the data the robot’s receiving is correct data for one, and we’re not, what we call a squirrel, chasing things
that aren’t there. And the paths are a lot cleaner. Reliability’s there. You’re not relying on mill scale or the cleanliness of a joint. As long as the weld joint’s
visible in field-of-view, we’re able to track it and decrease those cycle times, and in a lot of cases
get better products. – Tell me about
optical seam tracking from a usability standpoint compared to touch sensing
or through the arc. – Yeah, so this is a question that gets brought up quite a bit when we start
talking to new users. A lot of newer users are
afraid to get into robotics, let alone optical seam tracking
because of the technology. But the thing that
I like to point out is technology in general
is getting better and easier to use. I mean people find it easier to acclimate to a
touch screen phone. You find people are
more easily acclimated to a computer and
things like that. Well, industrial technology
is gaining traction in that same way. With our interface,
with the TH6, we have a really nice,
easy-to-use GUI app that’s very well laid out. The calibration setup with
Binzel has been very efficient. We have a very simple
calibration that we have to run. The setup itself is all
ethernet connections. And the TH6 viewer
integrates really well with what we’re currently
doing on RobotStudio. So it’s paired
together really well. – With the ease of use, the TH6D really comes
in pre-programmed for all the basic weld joints. You don’t have to build
your own profile up. You select one of many
that’s in there already, and they come with every system. And all of our
models that we have all connect to all
the ABB robots, from even some of the older
ones with serial communications. So we can go from
little older robots all the way up to
the brand new ones with good communication, good setup, and as Matt said,
the calibration. It’s simple, it’s
easy, it’s reliable. One thing I do like about
the ABB product line is they’re able to take and run our torch packages and our setup and use multiple swan necks. So if you got different TCPs, sometimes you talk about access when we got into
access in the parts. You’ve got this big
chunk of steel out there. But really I need to
get a longer swan neck so I can drop off
the whole package, grab a new sensor/torch
combination, and get into that access. But that one might be
different for this part. So there’s a lot of flexibility that ABB brings to
the weld process and being able to get a true
solution into the industry. I think with the Binzel torches, Scansonic sensors,
and ABB robots, you’re gonna find that the solution’s
gonna be there for you. – Can you guys go into a
little bit more detail? Tell me about the data
the sensor’s collecting and then tell me what the
robot does with that data. – We have three optical lines. And what that allows us to do is we’re able to collect
data not only in an XYZ. We can give you the
deltas around the XYZ. We’re collecting gap, mismatch, area is some cases. We have a plethora of data
coming into our sensor and we transmit all that
data back to the robot. ABB does a great job of
taking that information, applying it to
what they’re doing. They can do some stuff
with the adaptive. Matt, can you describe what, when we give you
this information, what are you guys doing with it? – There’s so much
data coming to us that we have availability
to all of that. So as Scott was saying, with gap just to start with, that’s an issue
that we see a lot, that there’s a gap in the part. We can start controlling process and the robotic
movement from there. If your gap starts to grow, we can do a couple of things. We can slow down
the robot movement, so reduce the travel speed, or we can increase
the wire feed speed to put more material
into the part. Or we can start also
adjusting weave data. Or we can do a
combination of all three. So there’s a lot of availability
to adjust on the fly with the amount of data that we’re getting
back from the TH6. – Why would you recommend
optical seam tracking over other seam
tracking solutions, or why would you recommend it for manufacturers
who come to ABB? – Yeah, absolutely. We actually look at it from
all ends of the spectrum. So from somebody that’s
buying their first robot cell all the way up to an automotive or tier supplier for automotive. For the simple fact
of the ease of use, ’cause that has always been the, the stopping point per se, is “We don’t want
to get into this “’cause it is very
complicated and difficult.” Well, nowadays with technology
and the development, it has made it very easy to use. – Okay, well that was a
lot of great information. I want thank you both so
much for your time. If you guys have any questions about automated
welding solutions, feel free to
contact ABB Robotics or Abicor Binzel for
your torch packages, sensors, robot periphery . This is Matt Sciannella with
Matt Conklin and Scott Huber. Thanks for watching, everyone. Have a great day.

Leave a Reply

Your email address will not be published. Required fields are marked *