All You Need to Know About Optical Seam Tracking
Optical Seam Tracking for robotic welding and hard automation have been available for years. Today, these devices are more advanced and productive than ever, as they help robotic welding operations improve throughput, reduce scrap, and drive
productivity. In turn, these solutions have become more user-friendly and easier to implement into welding cells. But, not every seam tracking solution out there is right for you. So, how do you know? In this free guide, we dive deep into Seam Tracking options, it's features,
and the best fit applications for Seam Tracking to guide you along. If you are an Automation Engineer, Robotics Technician, or Production Manager, you'll want to read this guide and learn about the ways Seam Tracking can impact your automation operation.
Want the Printer Friendly PDF? Click the Button Below to Download:
Ready to Start?
Scroll Down to Read All You Need to Know About Optical Seam Tracking. Enjoy!
Table of Contents
The most Frequently Asked Questions on Optical Seam Tracking answered
About the Authors
Scott Huber is Key Account Manager at ABICOR BINZEL with a special focus on seam tracking and laser applications. Scott is a graduate of Ferris State University’s Welding Engineering program. Scott leads the implementation of automated solutions for BINZEL clients across the United States and Canada. His primary focus is implementing optical seam tracking solutions designed to reduce cycle time and modify weld tooling for hard and soft automation processes. Scott can be reached at email@example.com.
Tom Graham is Key Accounts Group Manager at ABICOR BINZEL. A graduate of Ohio State University’s Welding Engineering program, Tom specializes in arc and laser welding processes and automated solutions. He leads the Key Accounts Group for BINZEL supporting key clients across industries. Tom has authored several white papers and technical articles on laser welding processes. Tom can be reached at firstname.lastname@example.org.
Is This a Familiar Issue?
Bob is the shop foreman for a mid-sized widget manufacturer in the mid-South. He walks into the shop at just past 6, pours some coffee and checks the scrap heap. It’s the first thing he’s been checking for the past month.
Why’s Bob checking out the scrap heap? Production has been having big problems with scrap lately. He sees from the second shift runs going into first shift the scrap heap is filling up. Production just can’t get consistency out of their automation process. There’s variance from part to part, lot to lot, the tools are wearing out far faster than they should, and the shifts are just not getting enough repeatability to the robot tool center point.
It’s been happening for weeks. Management is pressuring Bob about the high scrap costs, telling him they need less waste and less rework. Throughput is being affected; they have to make so many parts a day and they are spending more time fixing parts than what’s coming off the lot clean. Bob needs to figure out how to get more parts of this robot cell in a lower amount of time with a higher level of quality.
So what’s out there?
Complex Problems with Few Solutions
Bob goes to the manufacturing engineering team and asks them, is there anything you guys can look at that can stabilize this process? There isn’t enough budget to spend on tooling; we need a lower cost more permanent solution. The manufacturing team looks at solutions out there and optical seam tracking looks the most attractive, least intrusive to try out. So they fit a system on a cell and test it for a couple weeks.
The results are staggering. Scrap goes down 50%, the cell produces better welds, and the shift gets better throughput. They try a second cell, and now they’re beating production goals, decreasing lead times, reducing cost, and looking to expand their vendor base.
Common Challenges Answered for Reducing Scrap and Rework
The challenges of Bob are the challenges of thousands of shop foremen and engineering teams every day in manufacturing. And it’s not just limited to robot cells. How do I conquer the challenge of part variance when the customer demands perfection and my tooling can’t meet that demand enough?
It’s a problem with a solution. Many of you have read and seen optical seam tracking solutions. Maybe even considered it before, but didn’t think it would pay itself back quickly enough or make that much of a difference in your production shifts. But it has for many manufacturing teams in North America and is still the best option for overcoming too much scrap and rework in your production runs, among myriad other obstacles like part variation, etc.
When first introduced, seam tracking was a solution for the truly technical minds in manufacturing. Today, it is far more accessible. Maybe you knew about it and forgot, or aren’t aware of the advances it has made. This e-book will shed light on what seam tracking was, what it is today, and how it can be used in ways you maybe never thought possible and can be done with far less programming or technical know-how than you probably ever imagined.
It’s not as simple a question as it seems. Sure, seam tracking has been around for over 20 years now, but how much do you actually know about it? What do you think it does? And why is it beneficial?
This section will give you a basic introduction to seam tracking, cover some of the histories of it, and what the benefit of it is exactly.
We’ll also touch on how seam tracking works, what goes into it, and what the user gets from seam tracking that actually helps their production operation.
Optical Seam Tracking: Basic Introduction
The most basic introduction to optical (or laser) seam tracking is that it’s an added technology put on your robot that allows the robot path to be modified in real-time to better match up with the part and position relative to the program path.
That’s a lot to digest so let’s simplify: Optical seam tracking provides the robot a stream of data that it utilizes to offset the programmed path so the robot arm welds where it needs to weld and not where it was programmed to go.
In short, seam tracking eliminates error and accommodates for part to part variation.
A Brief History of Seam Tracking
Seam tracking itself has been in the welding industry since the 1990s. It started with abilities to find the seam via “touch sensing” where a small amount of voltage was applied to the wire and when the robot touched the part, it provided a short circuit where the robot recognized this change and recorded this location to a data register.
The next evolution of tracking utilized arc characteristics to track a seam. Through-arc seam tracking, known as TAST, was to utilize voltage variations induced from arclength changes seen during the weaving of the torch and resulting arc along the welding path. As the systems sees a change in arc length/current, the robot was able to modify its path to stay in the seam. This type of tracking is typically limited to fillet type welds where a weave could be utilized. It provided and still today provides simplification of programming for those seeking to correct joint misalignment.
Through-arc processes cut down on cycle time and rework. At the time they were revolutionary welding technology but didn’t always solve the problem of welding automation because of issues with the level of quality.
Vision-based, or optical seam tracking, is a more complex solution than through-arc seam tracking, but is regarded as faster and more reliable, and was first introduced in the late 1990s. Since its inception, vast improvements have been made in the filtering of the data, the speed of information exchange, and the overall ease of use for these types of devices, which has coincided with the advancements of computer programs.
Optical seam tracking as process improvement is a means for allowing your welding process to be:
- More efficient
- Achieve increased throughput
- Attain higher first-pass quality
Provide a lower overall cost in the long run for your automated processes. Consider the problem and the shortcomings of your robotic welding cell using older solutions like the through arc seam tracking.
Major points of variation in the welding process come from inherent variation in single parts. Take a look at your automated process. When you go from lot to lot or from run to run, one can expect there to be changes in material properties (elongation, spring back after forming, etc.). While that may be an inspection situation, the issue still may be outside of the process capability and require additional set up time or tuning at the changeover.
Robots alone don’t accommodate for these variations because the part isn’t always where the process is. Laser seam tracking as an adaptive tool gets your process to match up with the joint location.
So, what do you get out of optical seam tracking?
In short, you get a lot of benefits when used to its fullest potential. Properly deployed, optical seam tracking will lower your heat input and eliminate over-welding, reduce possible metallurgical issues, and allow your automated process to run faster and much more efficient. Without question, it produces a better-quality product at the end by offering the robot a means of adaptation and an agile solution to your welding process.
More Information on ABICOR BINZEL Seam Tracking Solutions:
Browse ABICOR BINZEL Seam Trackers:
First, there’s the seam tracking sensor itself. The main part of the seam tracker is a sensor package consisting of the sensor, communication/power cables, air hoses for cooling/cross jet, and system controller (similar to a small PC).
The controller is typically integrated to a robotic or gantry-based system with a multi-axis slide, sends the positional data to the device’s motion control package, and directs it to deviate from a programmed path. This deviation provides improved path accuracy to the joint because it is tied directly to that particular set up and not a general path that was programmed to another set of parts that wasn’t designed to account for variation.
Usually, with optical seam tracking packages, there will be an Ethernet or analog-based interface you would use for settings parameters of the system. These parameters will include joint type, lighting and aperture configurations, and seam profile (lap, fillet, etc.), so that the system can formulate the best tracking conditions of the seam in question.
The intensity of the laser light and aperture speed can be tuned to provide the best output display of the seam. Filters can be put in place to block out parts of the field of view that are not pertinent to the joint being processed. Blocking out parts of the field view is a particularly important step to take in cases where multiple seams are in close proximity to one another and only one is being chosen.
After tuning a seam tracker, which usually takes only a few lines of program data depending on the seam tracker device, the system and sensors start to do the work.
With optical seam tracking sensors, a laser beam is emitted from a diode that is either integrated into the unit or is part of a liner generator package. The beam is then converted by an optical line generator into one or multiple lines depending on the make of the sensor. The laser line then projects from the diode unit and hits the measurement object. It then diffusely reflects at an angle such that it hits a cloud of individual, more or less bright points of light, onto the camera chip. The individual pixels are then filtered and summarized into an ASIC chip.
Filters, both software and hardware related, will remove the reflections and other light influences. The data is then calculated by the seam tracking controller, and the position of the points of light on the chip are converted to positional information as it relates to the process. A profile of the joint is assembled and the results are linearized and compared with those of the user-chosen seam profile.
Other supporting parameters are then calculated in real-time with respect to the profile and compared with the defined values of the parameters set within the motion control side of the process. Offset data is determined and then output to the machine control, which results in a modified robot motion path according to the offsets found during the tracking process. This eliminates the tiny operator and robot programming errors that plague welding automation processes.
The information output from the sensor package will include information like X, Y, and Z positions, gap, mismatch, and angles (roll, pitch, and yaw) about the sensor datum. Analysis of the data provided can indicate at what points of the process greater adjustments are being made to accommodate the variation in the parts. This information can then be used to determine not only how the current product can be optimized, but also show where future product designs can improve.
Maybe you’ve seen articles and videos of seam tracking, but you’re not sure if it’s right for you.
It’s important to note that Optical Seam Tracking has boundless applications in manufacturing and production and can tackle some of the most difficult parts of your production - the ones that cause you headaches all the time.
We’ll dive in and explain how in this section.
In Which Case Would Optical Seam Tracking be Right for Me?
Optical seam tracking fits virtually any process that requires adaptive technology to adjust to variation in the workpiece where robotic or various semi-automatic applications are used.
In cases where there is part to part variation that is greater than the repeatability of the programmed path of the machine, alternative technologies are required to ensure the machine to part interaction is proper. These adaptive technologies include vision, tactile seam tracking, optical seam tracking, or other sensor means to identify where the part is located - either in the machine, on the machine, or before the machine, that is within the machine’s workspace.
Robots go where you tell them to go. If the part isn’t there, then the process isn’t done properly. This leads to scrap, poor quality, revenue loss, man hour loss, etc. Seam tracking allows you to build quality into the process instead of relying on human programmer error. Optical seam tracking is a means for helping ensure the process is done right the first time.
In situations where the part is too large to fixture, there is rough tooling that typically has to be done. Large parts usually mean multiple weld joints or large weld joints. Seam tracking is great for applications like these because it reduces the programming complexities and increases throughput results.
Seam tracking can be used across various materials across various industries. Conceivably optical seam tracking can be used to perform sealing, drilling, material removal… it can help improve the accuracy of the process on a given part. It isn’t captive to a single industry or process.
Seam tracking can be used to provide adaptive motion control to the automation of products, thus ensuring part and process meet.
Despite a wide range of possible usage, seam tracking is not suitable to every process or application.
Part shape, size, accessibility, etc. play a role in determining whether seam tracking is a good technology solution for your automation needs. The age of your current equipment is another factor.
In most current technologies there are ways seam tracking can be applied. In welding, butted joints and highly reflective materials are two application challenges that have been troublesome for seam tracking for years because of the inability of the sensor to handle large amounts of diffracted light along with. But today, with the right optical seam tracking device, that problem does not exist anymore.
What Equipment Does Seam Tracking Work With?
Optical seam tracking is ideally designed for robotic applications. Major robot makers like Fanuc®, Kuka®, Motoman®, ABB®, etc. integrate optical seam tracking via a sensor interface package held on the robot controller. Newer robots will lean towards an Ethernet-based interface, which is faster and easier to use than analog. Analog interface is more common for older robots or hard automation solutions like gantries, etc. Depending on what your equipment is fitted out with and how critical your cycle time is would dictate whether a newer, Ethernet-based solution is needed or an analog-based solution would work.
For mounting a seam tracking device onto the robot, the focal point is based around the tool center point of the robot. It typically leads the robot such that it collects data, sends the information back of the robot controller, and offsets the robot motion to match the true part position against the programmed path of the robot.
Sub-arc, tank welding, gantry and rail components with a single or dual-axis positioning are other applications that can use optical seam tracking to great effect. In these cases, the sensor integrates to the control via an analog base, as most of these controls aren’t set up via Ethernet nor do they have the motion control protocols needed to process path adjustment. The analog will then give off a voltage-based output to guide the motors and instruct them to deviate in one direction or another based on the data collected.
Think of optical seam tracking in hard automation like an operator with a joystick driving a carriage up or down to accommodate variations in the process. He sees it, then he nudges it in one direction or another to anticipate that variation. The sensor does the same thing. It processes it, outputs the response to a controller, and sends voltage to the motor to offset that path.
Beyond MIG welding applications, seam tracking can also work with plasma welding, plasma cutting, TIG welding, laser welding, laser cutting, grinding, and other various assembly applications. You can also use it in some forms to perform sealant. A lot of these applications depend on if the toolset can integrate a sensor, and whether the robot is programmed to accommodate a technology like seam tracking as part of its process.
There are three main elements most would be looking to seam track:
- Weld joints
- Material (specifically metals)
In terms of joints, there are 9 main joint types most optical seam tracking system can track. However, for a lot of seam tracking systems, these joints will not be pre-engineered into the system. You will have to take the time to pre-engineer these weld joints or get your technical support to do it.
Any optical seam tracking system that comes with the joints pre-engineered represents an advantage over other solutions out there. Some products require the welding operators to scan the weld joint prior to programming and defining joint targets. That can be time-consuming and increase the possibility of user error, especially for those new to the technology.
These 9 joints are the dominant joint types out there an optical seam tracking system will be able to track without issue depending on materials:
- Tailor Bank
- Lap Joint
- Open Side
For materials, the best seam trackers can track anything as non-reflective as black oxide and mill-scale mild steel to galvanized to galvannealed and stainless. There are seam tracking systems available that can even track extremely reflective materials such as aluminum and even diamond plate.
Seam tracking system that can track reflective metals is a pretty unique offering if it can be seamed tracked reliably. For manufacturers who use diamond plate and aluminum, for instance, there never before has been a seam tracking system that would reliably track a joint because the reflection of the material made the sensor’s sampling inconsistent and unreliable.
Today, with advanced sensor technology being incorporated into these seam tracking devices, there are now options out there able to track these kinds of materials.
Exclusive with some seam tracking system are the advantages and abilities of a three-beam seam sampling system. Such a system has many benefits, including:
- Ability to calculate deltas around X, Y, and Z-axis plain. With a 3-beam sensor, the sensor provides Y, Z, gap, area, mismatch, and length of the laser line. Also, available data are angles (roll, pitch, and yaw) about the sensor datum. Some robot OEMs take advantage of this available data while others do not. It should be noted that with some robots, the full advantages of 3 lines cannot be realized.
- Redundancy of track data information, which in turn gives you the ability to track those reflective materials and reduces the likelihood of the system being compromised.
- Angle detection of the part relative to the angle of the camera. This development allows for more accurate TCP positioning. Makes the TCP calculated and known through one scan cycle as opposed to multiple scan passes and actually having to formulate the TCP based on single beam seam tracking solutions.
The big difference in taking and using data from 3 lines of seam track technology is the ability to average out the measurements from the 3 lines. Pitch, roll, and yaw is 3-dimensional data around the TCP so that data rotationally around the wire can be captured to help for more accurate calibration of the robot to the weld joint. What is common to all robot interfaces is the ability to utilize the average of the 3 lines; thereby maintaining a 3-beam advantage.
More Information on ABICOR BINZEL Seam Tracking Solutions:
Browse ABICOR BINZEL Seam Trackers:
There are several seam tracking options out there, and not all are made or programmed the same.
There are different concepts and intent behind every seam tracking product, and you’ll have to sift through those options to determine which may be best for you. Here are some of the factors that are worth looking at when you’re making that decision.
One of the things that will separate seam tracking options out there for any person interested in using them is simplicity.
When you integrate technology like optical seam tracking, you want it to be easy to use.
The demise of seam tracking is complexity. When looking at the options out there, look at what has to be pre-trained.
There are typically 9 joints you’ll be using if you utilize seam tracking in a welding application.
Once you are left with the system, do you want to take the time to pre-engineer all those joints?
There are seam tracking options that pre-engineer these joints into their program to make integrating into your process far more easy.
How Much Does the System Sample?
This is all about sample size and reliable data. Most systems do one sample per scan. That scan is typically limited solely to the geometry of the seam in question. Superior seam tracking systems will do not just the geometry but also track the shadowing of the material.
Shadowing is significant because if you’re doing two dissimilar jobs without a joint or seam to track, there is little geometry to see. If one side of the metal is light and the other side is dark, the beam can track based on the shade of the metal.
If you only go off geometry, shiny metal can become very difficult to track because of shadows being cast or not being cast, which would limit the effectiveness of the seam tracking.
What's the Image Processor?
The biggest difference in seam tracking devices from a hardware standpoint is the reliability of the image of the picture taken from the scan. Images processed using a CMOS sensor array, as opposed to a CCD or PSD processor will generally produce better and more reliable results.
PSD and CCD – short for Position Sensing Detectors and Charged Coupled Devices – are two different breeds of sensors. PSD is a purely analog solution with a limited scope that primarily measures the center of gravity in a light spot.
It is accurate, but does not do anything beyond its base measurement. It can’t, for instance, differentiate a direct beam from a reflected beam. CCD sensors record light as an electric charge within each diode. CCD has shortcomings, however, in that it’s dynamic range is limited and any sudden shift in light intensity can cause it to falter.
So what makes CMOS superior? It’s a more developed technology with greater innovation. CMOS sensors work at a higher speed, cost less, filter out noise, and consume less power. CMOS works using exposure bracketing so it can accommodate low-light environments. With CCD, getting the data off the sensor can be more difficult because there are less analog readout channels, and the analogto-digital conversion takes place outside of the sensor itself. They also tend to run hotter, which can increase the likelihood of damage to the device.
CMOS, short for Complementary Metal Oxide Semiconductor, processes these small variables in the weld up more effectively because it’s technology is better suited to filter out “noise” from outside influences during the seam track. Outside influences can be anything from light emanating from the welding process or reflective energies coming off reflective surfaces, changing material shapes, line marks, scratches, etc. Welding process can be anything from spatter, the UV light off the arc, smoke, etc.
Having a separate processor outside of the sensor is another important hardware factor to consider. Users generally regard having a separate processor as an advantage because of the replacement cost. If the robot crashes, for instance, and breaks the sensor, or if you were to forget to put the cover lens on and damage the optics, the cost of replacing just the sensor is far less than replacing the processor and the sensor itself.
Another factor to consider when comparing seam tracking devices is standoff distance. Short standoff distances have to be more in tune with clearance issues. The longer a standoff distance in an optical seam tracker, the tighter the space it can access and the less it worries about clearance issues. A higher standoff distance also lessens the chance the device is compromised by weld spatter because it is positioned farther away from the weld seam. Look ahead distance is another separating feature.
Look ahead distance, in particular, speaks to tighter radius and response. The closer the track is to the arc, the more accurate you will track the seam. Diameter is another impediment. If you’re doing tubing or pipe, for instance, the shorter look ahead distance allows you to track smaller diameter piping. The track path likewise is better.
Between various seam tracking sensors maintenance is fairly standard. General maintenance comes down to the frequency of cleaning or replacing the protective lenses. The protective lens is a replaceable item typically located between the actual optics of the sensor and an air knife/cross jet which blows air parallel to the sensor to eliminate smoke from the viewing path of the sensor and help knock down any spatter coming from the welding arc. Maintenance will have a lot to do with your process. Considerations will include how much smoke the welding cycle emits that comes into contact with the lens, and the amount of spatter that can build on the lens of the seam tracker.
The tendency with seam tracking sensors is for production lines to wait until the sensor lens fails or is failing before initiating a replacement of the lens. Often times seam tracking failures will occur through degradation of the scan, or when inspection and replacement of the lens are not considered as a critical or routine part of the welding maintenance schedule.
A lack of understanding regarding the importance of the lens’ function is another usual suspect for seam tracker maintenance issues. When the lens needs to be replaced, the sensor cannot be expected to function without one. The reason being is the lens protects the camera and laser aperture. And if either of these aperture’s gets damaged because it’s coming in direct contact with contaminants (i.e. spatter or flux), the maintenance becomes a far more expensive and time-intensive issue that often cannot be remedied on-site. Why does this happen? Usually, it’s because of new personnel, lack of training, or an inconsistent or ignored routine maintenance schedule.
Maintenance of a seam tracking sensor comes down to knowledge and discipline. Think of maintaining the lens of a seam tracker similarly to maintaining a torch through a cleaning station. It’s logical to estimate the lens of a seam tracker will wear at a pace similar to that of one of the components of your welding torch – be that your nozzle, your diffusor, or your swanneck. Build lens maintenance of a seam tracker into a maintenance procedure like any other welding tool, and it will prove reliable and durable in production.
Another maintenance item with seam trackers to keep in mind is the hardware’s temperature. Seam trackers that keep all of their electronics inside the sensor tend to run hotter than those with an external processor
Add to that all the heat from the welding torch itself along with ambient plant temperature (hot summer days). Typically, you will need to add a chilled air-line into the backport of the sensor to keep it cool while in operation.
All sensors require a low-pressure airflow which serves as an air knife; blowing air near the lenses to help keep contaminates away while not interfering with the shield gas. Another item to consider or look at when installing and implementing a seam tracker is a filter on any compressed air line you introduce. Why is a filter needed? Typically, it’s not feasible to dedicate an entire line of air to the sensor, and sometimes air lines can become contaminated with oil or moisture can get into the sensor hardware and cause it to fail.
Compressed air line filters are thus small but required components to have for a seam tracking sensor to maintain the air knife function and help keep the temperature-controlled during operation.
For an application that would add significant heat to the hardware, a waterline can also be utilized that runs through a backer plate attached to the sensor to help keep the electronics cool while in operation. Backer plates are normally utilized in chilled air lines, as well. Seam trackers that experience less heat in the hardware will typically be ones whose processor isn’t builtin to the sensor but contained in a separate enclosure connected via a port; rarely will these ever need to use a waterline and, in many cases, can rely on an air knife to keep the sensor adequately cool.
Software updates are another maintenance item to keep up to date with. Always stay in touch with your seam tracking point of contact to get kept up to date on maintenance items in the seam tracking software. Such updates are usually made on a case by case basis.
Sensors must be mounted ahead of the welding torch to provide correctional data about the seam. Since there are a myriad of torch types and even robot arms, there can be no standard “one size fits all” bracket. The bracket design must be such that it provides accurate, and stable, positioning of the sensor to the robot TCP. Anytime the seam tracker moves as a result of the heavy torquing of the robot the tracking of the seam will be compromised as a result.
A solid, stable mount – like anything else on the robot – is crucial to a reliable tracking of the seam. This consideration is more commonly seen on over-arm robot torches than through-arm. With mounting to seam trackers, having a standard bracket offering in place makes accommodating a seam tracker much easier. For seam tracking providers who have torch offerings as part of their product portfolio, this tends to be a far easier endeavor than for providers without, because mounting brackets can be designed and standardized for specific torch packages ahead of time as opposed to having to design for third party torch packages in one-off scenarios.
Most of these one-off type brackets will also be designed by the integrator or OEM in such an instance and not by the seam tracker maker which can make mounting multiple seam trackers on multiple robots with multiple torch packages less standardized across the line.
Access to seam trackers is likewise an important consideration with mounting. With seam tracking, the most restriction will come from fixtures that are already in place and the process the end-user has designed into the robot – such as when the torch is already programmed to reach into a tight weldment or another part may block access. Most times robots are programmed with enough clearance to fit a seam tracker but there are situations where seam tracker access is not possible under the current conditions.
Another challenge or consideration to mounting with seam trackers is whether the seam tracker is being introduced to a new or existing welding process. Existing processes can sometimes be tight fits; oftentimes it’s just designed for the torch to access and another component like a seam tracker was never designed or intended for the process. Situations like this can make for challenging mounting obstacles for seam tracking, as redesigning tooling and clamping as well as clearance to fit seam tracking into the process would have to be weighed against the investment of having to redesign some aspects of the weld process, like fixture design, or modifying the robot welding procedures which could affect critical cycle time.
With seam tracking, it’s critical for the mounting of the seam tracker to fit the process. It’s difficult for a welding engineer to change the torch angle, length, etc. to fit seam tracking into the process. As such it is incumbent on your seam tracking provider to have the right type mounting that can fit the process and see the joint accurately. Whether the torch is fitted on an over-arm or through-arm robot the mounting challenges carry the same degree of complexity.
With hard automation, when using sub-arc, there are different sets of considerations to make.
One is that there is typically no spatter or as harsh a light thanks to the flux; so, the maintenance requirements (relating to lens cleaning) are substantially less.
However, the consideration with sub-arc and seam tracking is that you must avoid the flux coming into the path of the sensor which would cause faulty measurements.
To avoid this, designing a gate or deflector dam for your seam tracker mount that blocks the flux from getting into the sensor path is sometimes a necessary consideration.
Hard automation will see more custom bracketing challenges than with articulated automation because of the custom gantry or articulating arm systems you’ll see more often in hard automation. Two, sometimes three-axis slides, will also be used more as means of mounting the seam tracker and torch. One frequent question asked about seam trackers is how much do they weigh and can a slide accommodate it? The answer is adding a seam tracker doesn’t usually overload a set of slides.
Now that we’ve learned about the things that separate Seam Tracking systems, how Seam Tracking can be integrated into your automation, and exactly what Seam Tracking does as a process, can you use it?
It’s actually a lot easier to use than the process you currently use. Seam Tracking’s stated goal is to uncomplicate the difficult parts of automated production.
Seam tracking is user-friendly, especially when you consider the alternatives. For a quality seam track, in some cases, the number of programming points in the weld path can be as few as 2 programming points. The number of programming points depends on the complexity of the part. Seam tracking can work on parts when they are either flat and stationary or when integrated with rotating or moving positioners. Consider if you were to try and weld a rounded shape with just a robot arm. You would create infinitely more programming points, and not achieve the same repeatability or quality. Optical seam tracking suits such a welding process perfectly.
Ideally, you’d want your optical seam tracker software interface to be pre-engineered with the joints you’ll want to track. That way all you’re looking to do is make slight adjustments to the existing engineered seams, program the points on the robot controller, and go. It makes communication with the robot far easier.
Spending the time and energy to develop those joint algorithms into your sensor will likely mean you’ll need to rely on technical support from the vendor to get you up and running, with a lot of phone calls and back and forth in between while you are perfecting it’s integration into your automated solution.
Do all of the optical seam tracking systems track with such repeatability using such a low number of programming points?
Do they all come with the pre-engineered weld joints?
Do your research and ask questions, and look at as many options as you can in person.
If possible, ask if they can show you their system against your process so you can see the benefits and the ease of use.
We've touched on a multitude of subjects regarding Optical Seam Tracking. While the goal is to keep this review of Seam Tracking as objective as possible, there are several frequently asked questions about Optical Seam Tracking Sensors and the Scansonic TH6D in particular. This section is devoted to answering those questions in as much detail as possible.
It works on the concept of laser triangulation. A laser diode inside the sensor creates lines and projects to the part. These lines reflect and are picked up by a camera. Data is then turned into lines which the system interprets as the joint and sends positional information to the motion system. Information output includes data for X, Y and Z positions, rotation angle around each axis, joint, gap, mismatch, etc.
Three times the sampling allowing for excellent stability during the acquisition of measurement data. The sensors use the average of the three beams to calculate the path allowing for greater variances in material quality:
|Single Line Seam Tracking Solutions||Three Line Seam Tracking Solutions|
|Seam geometry information||Y, Z, gap, area, mismatch, length of the laser line||Y, Z, gap, area, mismatch, length of the laser line|
|Angles||Not available, theoretical about comparing two or more pictures and robot information||A, B, C right, C left, C total|
|Measure redundancy||None||In every picture|
|Plausibility||None because of the missing redundancy||Yes because of the redundancy|
|3D Applications||Only possible with additional software or multiple part measurement||Possible because of all the necessary information is available in one picture|
|Equivocality||None||Equivocality possible; software parameters and filter supporting|
|Arithmetic average||Not available, theoretical possible calculating robot position with sensor measurements||Available|
Concave-convex part detection
|Not possible||Available due to the distance in "X" of the laser line|
The standoff for most Optical Seam Tracking sensors vary. For the Scansonic TH6D sensor, the standoff is 150 mm (+/- 12 mm). Look ahead in robotic applications is anywhere from 10 mm to 30 mm.
In Fixed Automation, the look-ahead distance is 10mm on the low side as far forward as the process will allow.
Yes, but there can be a little more setup as the sensor angle to the joint is not as forgiving.
TH6D Optical Seam Tracking systems use a data filter that is implemented by software and its purpose is to erase the reflections and all other light influences like the ones shown below:
- Welding spatter
- Excessive reflection due to highly reflective surfaces like aluminum
6 m/min (256 in/min) has been verified as possible with the right set up.
Actual speed will vary to some extent.
Most high-end arc applications are at 80 - 120 in/min.
The TH6D is capable of processing these speeds with good stability.
Robots will typically require a specific software set on the controller to support the installation and interpretation of data flow for motion control.
Most connections are done via Ethernet however digital/analog interfaces are available as well.
This unit can be used for virtually any process that requires real-time seam tracking including, but not limited to, MIG, TIG, plasma, laser, brazing, sealing, caulking, cutting, etc.
Hard automation is another great use for Optical Seam Tracking. Integrated hard automation can include a seam tracking sensor, torch mount, and 2 axis motor package to control torch position (Y and Z up to 200 mm total in each direction) based on the seam tracking information.
This can be used for gantry or fixed position welding and contains its own control package for setting up the seam tracking unit.
Frederick, MD 21703Google Maps001-301-846-4196