Stop Missing Splices: Automated Detection for Converting Lines | 1DC Sensor

Hello everyone, Arvvin Shadri from Roll for Roll Technologies. Today we're going to show a little bit about our splice detection system. Right now we have the 1DC sensor set up so that it can detect a splice. We just [music] saw a thing splice go through.

Whenever the splice is detected, we have a stack light and the output from the 1DC sensor is triggering a stack light. Apart from a splice, we can also detect voids or defects. You saw a big hole go through and a couple of labels go through right [music] now. So essentially, we can look at some surface defects by looking at the image we are getting from our 1DC sensor.

There's another hole coming through or tear was able to detect that. We have some [music] web instability like you see there. a little bit of a standing wave of wrinkles going through, but that is not being triggered. The sensor is able to distinguish between a splice and some web [music] instability and not trigger for that web instability.

At this point, we're running at about 200 ft per minute just to give you an idea of how the system [music] works and also for you to actually see the splice as it goes through. What we are going to do a little bit later is that [music] the same set of splices. We're going to run it at a much higher speed and see how that works. We do have different types of [music] splices.

Essentially, there's no setup. There's one condition it has been running on [music] and it can detect a hole, a bright splice, a dark splice, a transparent splice, and even a butt splice. So, we do have a transparent splice that is going to come up here. I'll point out when it comes up, but doesn't matter what it is.

Unlike other splice detection systems, we are able to detect all of these different types of splices without making any changes to our controller or the sensor configuration. our 1DC sensor which is 1DC 480 which has [music] about 7,600 pixels and all of this is running at about 200 [music] times a second. So even a bud slice or a really difficult to detect transparent splice, we're able to see [music] that. And sometimes it might be a high contrast splice like a pink material that we are seeing right here and we were able to see that.

The main challenge with splices is that the splices can have different colors. It can vary in the contrast difference between the material and the splice itself. And sometimes it's a hidden splice. That's what we're [music] going to see next.

We have a bud splice where we we have a double-sided tape that is going through right here. It came through. And [music] you have uh two layers of material on top of a double-sided tape. So, you don't even see the tape.

You see the material. It was able to pick that up. And just to give you an indication about different colors, we have a dark color splice coming through and it still picks it up. The other common thing that we get asked about is that how long of a splice tape can we detect?

So, for example, right now we're going to have a shorter tape, a darker shorter tape [music] that comes through. It's able to detect that. That's about a 1 in tape going at 200 ft per minute. We can also detect darker colors or tapes.

As long as there is a contrast difference between the material and the tape, we'll be able to detect that. That's what we want [music] to showcase. Here we have a darker tape coming through. As you can see, it can detect that, too.

Sometimes the [music] splice may not be on the right side. In this demonstration, we're going to show that as well. And obviously, a smaller splice tape, can we detect it? We'll look at tapes coming on different sides of the material and can we still detect that or not?

Another darker one came through. [music] We're able to detect that. While we run this, this is at a slower speed so that the user can actually [music] see the splice. Once this is done, we're going to run it at a much higher speed.

So, this particular sensor, ODC 480, like I mentioned, it's running at 200 hertz or 200 times a second. So, it's able to capture the image, process the image, and send an output within 4 milliseconds. If we have a smaller viewing area, it can run faster up to about thousand hertz of [music] processing speed with these sensors. We are taking that input, processing it [music] and sending an output through EtherCAT.

In this case, this output is connected to our PLC and then the [music] PLC is actually triggering the flag. Now, here's an example where the splice is not on the right side. If you caught it, you can see that the splice was on the other side of where that camera was looking at. Just to give you an example again, another one where the splice was on the other side and we're able to detect that as well.

There is a little bit of splice tape that was folded over and we're still able to pick that up even though the splice is not on the right side of it. just to highlight the versatility of our sensing system in order to be able to detect the different types of splices, different variations in contrast, the way that an operator would do that. Now, another transparent kind of splice coming through. Here we have a transparent tape that [music] is splicing together the material.

And as you saw, it was still able to pick it up and do it. Another one where the splice is on the wrong side. We're still able to pick it up. We can't guarantee that it'll pick up all the time, but most of the time we can still pick it up.

And that's what we are trying to demonstrate here. Another example of a splice. It's a transparent tape and then splice on different spaces. We're able to do all of these.

So, this is just an example of how our splice detection works. In the following, we will speed it up and run it at thousand feet per minute. So you can see how it's able to perform even at higher speed as long as we're able to see it. Again, another transparent splice coming through.

And we can also detect some holes as we saw before. We're able to see some holes and tears and essentially void detection and anomaly detection. All of these things we'll be able to do. So, let's now look at [music] it at a higher speed and see how it does.

Now, we ramped up,000 [music] ft per minute. Some 1 in gray tape went through. A 1in spring tape went through. A 2-in gray tape went through.

A 1-in gray tape went through a dark green splice, a budge slice, a pink tape, a transparent one. Now we are at the end of the roll, so we're going to slow down and then reverse direction again. And there was a label that went through and then a hole or a tear label went through. And then another pink tape.

Now the machine is going to pause and then it's going to go in the reverse direction. And we'll see the same thing in the other direction as well. Heat. Heat.

Here you go. all these places at full speed. So we were ramping up from anywhere from 500 feet per minute to 750 and the max speed was about,000 ft per minute, we were able to catch all the slices, [music] even the butt slice at 1,000 ft per minute. The main thing is that even if the contrast [music] difference between the material and the splice is low, this algorithm that we have for splice detection is able to catch that.

And best of all, you can use this not only for detecting splices but also tears in the web, holes in the web, voids in the web. And we'll have more videos talking in detail about this. Once again, Arvin Sashadri from Roll to Roll Technologies. Take a look at our 1DC sensor [music] for more information about how you can use it not only for edge detection, width measurement, but also defect detection like splice, void, stairs, and the foreign matter on the material.

Make sure to subscribe and enable the notification. Thank you.