Cycle Time with Robots. Faster or Slower?

Our follower Sean writes: I am a robotic weld engineer in the US. I am responsible for up to 97 robots. We have always talked about vision in welding but, the final cycle time will be reduced. Maybe, take a look at an auger housing of a residential snow thrower or riding lawnmower deck? Thanks.

Great question! Today let's jump into the subject of cycle time.

Understanding the Impact of Adaptation on Cycle Time

It’s best to start by saying that cycle time can get very complex. But we will show the general concept using simple diagrams. First, it is extremely important to consider the full production cycle, not just the welding on the robotic cell alone. Next, the answer heavily depends on the process characteristics of the particular production: batch size, - serial or high-mix - type of product (dimensions, complexity, et cetera), quality of part preparation, and quality of fixtures.

Let's see what will happen when we use robotic programs without any adaptation (meaning machine vision or touch sensing). In this case, we need to consider not only the initial programming of the part and cycle time of the welding process but also need to measure two additional procedures which usually take place when using robots without adaptation: touch-ups of the program after we load and unload a new part and rework if the touch-up has been skipped. The frequency and time required for touch-ups or rework depend on your particular case and should be measured to analyze the final cost.

What happens in the case of using robotic welding with adaptation? Firstly, the complexity of programming is increased. It also could happen that the programming of the adaptation is even more complex than the programming of the welding procedure alone. The cycle time of the robotic cell is slightly increased compared to the case without adaptation. On average, if you want to estimate adaptation time, it is roughly 15% to add on top of your weld time. But as a reward in this scenario, both touch-ups and rework are eliminated which can be very beneficial in terms of the full production cost.

If you’re using Abagy, an additional benefit is the simplicity of the initial programming which takes significantly less time compared to conventional ways of programming because the adaptation and all robot movements are calculated automatically. It will take minutes instead of hours. One more benefit is that the adaptation procedure provided by our software is not only automatically generated but also more advanced and can compensate for a wider range of deviations compared to the standard adaptation procedures.?

In other words, to understand whether it is beneficial for your particular case - to use adaptation techniques or not, you need to measure programming time (how often and how long), touch-ups time (again, how frequent and how long it takes), and rework (the cost of it).

Let’s discuss some typical use cases.

Typical Use-Cases: Small Serial Parts vs. Large Serial Parts

The first one to consider is the welding of small serial parts.?

The programming time is not the biggest concern as the batch size is huge. The touch-up procedure depends on the specifics of the part - the quality of preparation, whether or not it’s possible to create a high-quality fixture, etc. So, the complexity of the touch-up procedure and its frequency may vary. If touch-ups and rework are not a pain point for you, then there is no reason to increase the cycle time by using adaptation. So, Yes, Sean is absolutely right that Cycle Time can be longer.

Let's consider another situation. Suppose you have large parts, but the production is still serial, meaning you weld identical parts. In this case, adaptation becomes more efficient.

The large parts take much longer to program. It’s also much more likely that touch-up procedures will be required for almost each new part because it is not easy to fix the part in exactly the same position each time and higher deviations are typical for larger parts. So we can say that you will either have to spend a huge amount of time doing touch-ups every time or you will have a lot of rework. In this scenario, the adaptation is very efficient and will reduce the total cost of production.?

You should consider Abagy software in this case to reduce the time of programming the welding and adaptation procedures.

High-Mix Production: The Need for Powerful Adaptation

Now let's discuss high-mix production. After all, if you need to program each part, especially if it is a large part, programming could take more time than welding itself. Which is absolutely crazy. That’s why most of the manufacturers working with non-serial parts do not use robots at all.?

But let’s still compare robotics without and with adaptation for this scenario.

When you have a wide range of products to be welded on the robotic cell and the batch size is small, it is too costly to design and produce a custom fixture for each part. Thus you will end up with programming and touching up almost every part.

Manually programming an adaptation procedure in this scenario is not so beneficial because it will take even more time and it will not be highly-reusable, as the batch size is too small.

It means that in the case of high-mix production, you definitely need something much more powerful. This is exactly the case where Abagy software gives you the biggest value.

Vastly reducing the time of the programming and providing adaptability at no additional cost, this seems to be the perfect solution. You can see the comparison in the diagram above.

Comparing Robots and Humans in the Welding Process?

In the case of high-mix production, companies often don’t have any experience in robotics. And this is not surprising, because it has been simply impossible to use robots there without new technologies. Therefore, manufacturers want to compare not two programming methods, but a robot and a human.

Let's discuss this topic too. During the welding process, both the robot and the human perform three actions:

• Welding
• Adaptation
• Movement

Let's start with the welding. The welding speed is not significantly different between a human and a robot. There’s no getting around the physics of the process. Even if the robot is faster, it's not by much. You can achieve a time advantage of a few tenths of a percent, but not three or four times faster.

Adaptation. Humans adapt much faster. People just look at the part and immediately weld. In the case of autonomous robots, we need to scan the part using machine vision, then touch-sense the weld, and then recalculate the robot path with mathematical algorithms, matching the ideal 3D model of the part and the real physical part. But a human, unlike a robot, can be distracted, and spend a lot of time on preparatory actions. So, in an ideal theoretical case, a robot can adapt as fast as a human.

Movement. Here's where a robot can beat a human. This is especially true for large parts and long seams. It's inconvenient for a human to weld them. You constantly have to move, drag the welding machine along, and make stops during welding. In general, it's just plain inconvenient. A robot can win hands down. But this only applies to large parts.

Also, I should say the robot provides high and stable welding quality and can work 24/7. In the case of an acute shortage of welders, which is expected to reach 400,000 specialists in the US by 2024, robots are an excellent solution.

Analyzing the Snow Thrower Auger Housing and Riding Lawnmower Deck

Sean asked to take a look at the Auger Housing of a Residential Snow Thrower or the Deck of a Riding Lawnmower. Let's discuss the Snow Thrower Auger Housing.

First of all, to provide the most accurate answer, we need to learn more about your production process. Therefore, we can only make assumptions.

The Auger Housing usually is made of thin metal. It is very difficult and time-consuming to assemble the product while ensuring accuracy. In this case, the fixture can be massive and complex. This is mig/mag welding, which is what we specialize in. Most likely, because the parts are stamped from thin sheet metal, they are quite warped, and the pre-assembly is also crooked. In other words, the seams may shift, and gaps may occur.

As we discussed earlier, yes, Cycle Time can be reduced because we use adaptation (scanning the part and then using touch sensing). But if there are problems with defects, it is better to increase Cycle Time than to get defective products or to spend a lot of time for touch-ups.

Moreover, if you are using a fixture, we can reduce the adaptation time by automatically scanning and recognizing all fixtures. We will scan and touch the seam faster.

A riding Lawnmower Deck is also similar.

In conclusion, cycle time with robots can be either faster or slower, depending on the specific production characteristics and the use of adaptation techniques. The decision to use adaptation should be based on factors such as batch size, part complexity, and rework considerations. Abagy software offers advantages in reducing programming time and providing adaptability, making it a valuable tool in high-mix production scenarios.?

Sean, thanks again for your comment. All comments are welcome!?

Please subscribe to our Biweekly Newsletter.?

If you prefer video format, please watch our YouTube Vlog.

Follow us on?LinkedIn?to stay in the loop with all our latest updates and discoveries.

Best wishes,

ABAGY Robotic Systems?team.

abagy.com

#robotics?#robotprogramming?#automation?#manufacturing?#industrialrobot ?#welding?#abagy