ABIBLOG United States

Why TH6x's Value Proposition is Truly Unique

Written by Jason Jamiel | May 17, 2017 7:32:08 PM

When you're looking at the things to value in an optical seam tracking system there's a lot to consider. What joints do I need to track, can the sensor track it, how much time can it save me off my cycle time, how reliable does it track? Lots of question, sure, and maybe there are some others that you've considered as you've look around at optical seam tracking solutions out there.

But with seam tracking you have to get the most in terms of value, convenience, and effectiveness, and that's something TH6x can provide to the automated process in abundance.

With TH6x, which can include both the TH6D standard version and the TH6i for zero gap, the most compelling value it offers to the robot welding process comparison to other similar solutions is the shadow detection and geometry detection. This is specific to the TH6i, and speaks directly to the challenge of zero gap seam tracking. But not only does TH6i perfectly zero gap detection, it performs that joint track without a pre-scan, and a pre-scan is what most other vision sensors will do to reliably track the joint. Let me explain why that costs productions plants hundreds of hours in lost time every year: With the pre-scan, the sensor leaves the welding arc alone for the initial run and just scans the part.  Once it scans the part, which is done on something like a rotisserie, the sensors record the gap, then go back to the home position, then fire the arc, and track the weld in third person time. So while it does perform the job of tracking the joint, it's not doing so in real time. 

The reason most sensors perform in this way is because the shadowing arc light and the photonics - or the light value of the weld, are too bright to allow the cameras to detect zero gap detection. In consequence what you end up with is almost double your cycle time. If the weld takes 10 seconds and the scan takes 7 seconds, your cycle time becomes 17 seconds. When you're in the Tier 1 automotive, or any high production arena, you almost double your cycle time as a result, decrease your productivity, and lose value on the seam track solution. This is literally the case on dozens if not hundreds of robot and hard automation projects that happen every day. With triple beam sampling, like what the TH6i or TH6D does, the sensor instead detects the weld joint in real time, judges the face value of the geometry and the shadow of a zero gap - when necessary, and allows the robot to weld while seam tracking at the same time.

TH6D is a similar product, but it will not track zero gap detection as reliably as TH6i, which is what occurred on a recent project I performed. I could find the edge of the part and implement an offset, but you can't regulate the inch of the part the way a large series production operation would need because the measurement could vary, and if that length varies then so does the mathematical offset that you put in to dictate where that seam is. This is a method any optical tracker besides the TH6i wouldn't be able to perform, but the TH6D is very good for finding gap and wandered gap - which is the more standard way to track a joint.

 It's also important to note that TH6D is very easy to set up and use. Think about the technical know-how it takes to operate any automation device and it's software: Understanding and operating your equipment is such a key to optimizing it's use. What's unique about the TH6x as a a device is it provides standard pre-determined geometries built right into the software as a standard feature. The user interface is just as simple - an operator can be shown how to use the TH6x Seam Tracking sensor in 15 minutes normally - even if the operator doesn't possess a background in robotics or computer programming at a basic level. With other systems - and I can speak from my own personal experience - you have to be very well versed in robotics and computer coding in order to manipulate the data and understand the physics of what's going on with the tracking of the joint. TH6x doesn't require that level of expertise to use effectively.

Triple beam sampling as a feature is an especially unique technology feature of the TH6x. It's the only sensor in the industry that does that. The triple beam sampling process is easy to explain:  it basically is checking the joint 3 times at a very chronological point. Other sensor solutions will do detection once with a single beam.

So what's the tradeoff? Why is single beam a less effective method?

Simply put it's not as accurate. When I think about joint detection the thing I look at is what I'd call a "grab rate", or detection. An acceptable benchmark for any seam tracker is 40%. TH6D alone can detect at a 40% grab rate even at a zero gap - which it's not designed for, and in most applications picks the joint up at better than 95% detection, which means your tracking is 100% perfect with no issues tracking. The lower the detection rate, the less of the chance it tracks the joint accurately - but 40% is a benchmark application any optical seam tracking sensor should be able to meet. Using the TH6i for zero gap - which features the grey scale camera hardware TH6D doesn't have - you'll get a >90% detection even on reflective surfaces like a diamond plate aluminum provided you're tracking butt welds or lap joints.

These numbers have big consequences when you look at your production under a microscope. Think about for instance large series production like Tier 1 automotive projects. Ideally, you want to put a system in, set it and forget it, and start the process. The most you ever want to do is switch out your consumables and get to welding. You shouldn't have to go into the cell and touch up points if you're using a value add process like optical seam tracking. But no matter how advanced a robot you have, if your gap is wandering or if your part is wandering, and you don't have a seam tracking system accounting for that variables, you'll need to go into the cell and touch up points. And everytime you have to walk into the cage to do something like touch up points - it's downtime. Already when you walk into the robot cell to merely change a consumable, you're taking at best an 8 second runoff to switch a contact tip. If you have to do a touch up weld on top of that - or if you're throwing it to the scrap bin for rework later - it's another additional 5 to 10 minutes of lost time off your production. This result is so many wasted cycles that could have been optimized with a reliable sensor it defeats the purpose of having a seam tracker in the first place that is able to repeat the robot path over and over again while accounting for part variance. 

The other benefit of the right seam tracker is it's ability to keep the operator's hands off of the teach pendant. This makes maintenance's jobs easier, it makes tracking your welds easier, it makes understanding your welding history easier and more accurate, and it helps in keeping the statistics of your welds at a constant variable so you always know the kind of products your putting out and the quality of the product is reliable. What's better is that when you run your macros and you understand what processes you're using this part for and what it needs to do, you'll have nobody catching that robot or compromising it's output because the data stays in the system and isn't subject to any operator error or interference.

When it comes down to the cycle time of welding, considering more than the upfront investment in a solution like seam tracking is crucial - because the payback on these solutions lasts forever. If you have a higher investment solution that costs more, look at it's payoff. If the higher cost solution works like it should and within 6 months you paid off that extra amount it's win, win. Conversely, if you have a product that is extending your cycle time and you buy that solution because it's a lower investment you never catch up, because it will forever be clocking your cycle time and adding more time to the process, meaning less parts out the door at the end of the year.

In short, look at the options out there and actual put their performance to the test, and the right solution is going to clearly present itself to you in the end.