Robotic Welding Issues and Challenges

      Mitch Dupon
      Post by Mitch Dupon
      August 20, 2019
      Robotic Welding Issues and Challenges


      Decades of lack of focus and investment combined with a declining workforce has left the welding sector of the economy with a labor force challenge.

      There are not many things you can point to in your daily life that does not contain something welded. There is a shortage of welders in the market today. There will be a shortage well into the future. The overwhelming sentiment being proclaimed today is not a rosy picture of a strong and robust workforce rising to the challenge. Like most pictures, there is an element of truth in it. Decades of lack of focus and investment combined with a declining workforce have left the welding sector of the economy with a labor force challenge. This is the reason many of you don’t have time to sleep, can’t attend your kid’s ball games, can’t fish, can’t (insert your hobby here) as you strive to ensure you meet production quotas, maintain quality, stay competitive and grow your business.


      Two of the main challenges are the number of workers to fill open positions and the skill of the workers. With such an impact on our lives, the current condition of the trade makes you wonder what we have been doing for the last few decades. Well, the industry hasn’t been sitting back and sulking. The most common way to fight this problem is by automating the process.

      There are significant barriers to entry for automating the welding process and this has resulted in those advances being constrained to very narrow and well-defined sectors of the welding market. Through the use of the non-traditional, industrial robot, a collaborative robotic welding solution removes those barriers to entry and brings the benefits of automation into your shop. Not just the benefits afforded by traditional welding robots, but additional benefits as well. The automation becomes a tool in the hands of your current labor force, increasing their individual productivity, it does not become a standalone piece of machinery needing its own department within your company to keep it running and productive.

      Barriers to Entry - Utilization

      Taking advantage of higher quality requires precise fixturing of the parts that make up the final assembly. This takes time and money.  Job shops who receive orders of 5-100 parts can’t justify the expense of building high-quality tooling let alone automating the processes to manufacture the parts for the assemble. This is why automation salespeople start with the two questions: How many parts are you making? What is your cycle time per part? This helps them determine utilization for traditional robotic welding.

      Over the many years, automation companies have determined the combination of a number of parts/cycle time that will create converts. What I mean is this, if there is a high volume, they can demonstrate high utilization at minimal capital expenditure. They then move to weld samples to prove to you the numbers they have presented. In the automotive world, the automation buyers will determine the utilization themselves, as they already know they will automate but are determining how much automation to buy. In the large parts manufacturers (think CAT, New Holland, John Deer) utilization is essentially the same but is presented as an arc on time or deposition rate. The cycle time per part may be hours or days but the determining factor is how much metal the robot can deposit vs a human in a given time frame.

      Much of the large parts manufacturing requires additional technologies to accomplish their utilization targets. When compared to the repeatability of manual welding of the past, the automation of large parts produces a far more repeatable product. However, when compared to welding wire sizes, the weld joints still vary in space too much for the robot to weld in the same place in space every time. The industry has stepped up the challenge with technologies such as Touch Sensing, Through Arc Seam Tracking, line laser seam finding, line laser seam tracking during robotic welding. Each of these technologies adds complexity, cost, and time to the application of automation. The increased complexity means dedicated programmers that have weeks of training at $1000’s of dollars of cost, then need months of onsite experience to effectively apply and maintain the automation. The technology works wonderfully, if you have the utilization and budget you can build the tooling, buy the robot, add the additional technologies, train the people and be successful.

      If you have a High Mix / Low Volume production profile and push back on a salesperson who claims you lack utilization to justify the purchase, they will fall back to a couple of standard arguments. First, it will take you longer to program the part than it would take you to weld it manually. Second, any robotic welding system that you buy will have a limited capacity, both in three-dimensional space and in the ability to manipulate the part in that space. If you want a cell to handle everything you need a $200,000+ cell that comes with a big workspace and tilt to rotate capability. And this robotic welder cost doesn’t include tooling. For most that are priced out of their range. Anything less and you are sacrificing workspace or part manipulation.

      The barrier of entry – Cost of Ownership

      By cost of ownership, I am not referencing the cost of maintaining and repairing the automation. That is truly low when compared to maintaining an effective workforce. What I am referencing is the overall impact on a company. Every implementation of automation requires the establishment of an automation department. Whether it is as small and unofficial as the owner or family member taking ownership of the automation, getting trained, dedicating consistent time honing programming skills and keeping the production running, or a medium-sized company attracting and hiring skilled programmers with welding knowledge to take ownership of the automation, or a large company establishing an official automation department staffed with offline programming engineers, controls engineers and technicians trained in robotic welding. In each scenario, there are a limited number of knowledgeable, skilled individuals who can realize the benefits of automation for the company. The automation department becomes the axis upon which production orbits.

      In many cases, especially in remote areas, the company has swapped the problem of finding quality welders for the problem of finding quality robot programmers. Companies face the problem of employee retention. Welding robot programming knowledge does not come cheap. Once a company trains someone within their organization there will always be a competitor who will want to take advantage of the company’s investment by spreading the cost of training and development over time by paying a higher salary to the employee and stealing them away.

      Time and again traditional automation has attempted this feat. Square peg in a round hole, it has rarely been a success.

      Until recently there hasn’t been an answer to the question of how a company that has small batch runs and limited capital can bridge these barriers and solve their labor problems while simultaneously gaining the benefits listed above. Time and again traditional automation has attempted this feat. Square peg in a round hole, has rarely been a success. That is because these companies have a different set of requirements. They require a whole new approach to the problem.

      These companies are defined by:
      1. Low volume production
      2. High mix of parts manufactured at the same time
      3. Short delivery times
      Low Volume Production

      If these companies were asked two questions to determine utilization their answers would never fall into the appropriate ratios. The number of different parts would be high, the number of each individual part would be low. An example would be:

      A cylinder manufacturer that has:
      • 200 different sized cylinders that need end caps welded with fittings on each that vary in location between batches.
      • 20 of each part welded per batch with an indeterminate frequency. (when we get an order)

      With traditional automation the sheer complexity of creating hundreds of programs, maintaining the programs between batch runs, and relocating the program for the fittings' new location is untenable. It requires, either endless programming or complex sensors and software to auto-detect the cylinder size and fitting location and automatically make adjustments. For a company for whom the welding arc equals revenue, a plant manager watching the robot being programmed for a greater time than it is actually welding, it is like watching money burn for no return on investment. To apply sensors and software to automate the variance is expensive, sometimes requires experimental software but most importantly, almost always, is constrained to a tight set of parameters. Sensor protective covers need frequent changing, special lighting conditions need to be provided, base material limitations (aluminum is potentially too reflective as an example). More effort is expended understanding how to make the system work and then maintaining it, then is needed to weld the parts manually.

      With traditional robotic welders the sheer complexity of creating hundreds of programs, maintaining the programs between batch runs and relocating the program for a fittings new location is untenable.

      A tool that is easy to program and requires minimal training can bring the complexity of producing a small batch of parts into the realm of possibility for the Low Volume producer. When the robot can be simply led through its welding path, understanding complex welding software, coordinate systems, joysticks, plus-minus keys, and software commands, are no longer necessary. When the tool can be programmed from a phone or tablet with a simple app, your grandmother can do it. When training takes hours and not weeks, the system can be a tool that every single welder in your shop can use to be more productive. A part can be programmed in minutes and small batches of parts runoff in less time than the welder could weld them by hand. When combined with modular tooling systems, you have a combination that makes this a reality.

      High Mix of Parts Being Manufactured at The Same Time

      Most companies have many skilled welders manufacturing different parts. This presents a problem with traditional automation, though there is a lot of robotic welding taking place it is not all the same. A traditional automation cell will have one-two station. A flat table in front of the robot or a turntable with a side A and B are typical examples. This means a maximum of only two different parts can be welded at once. In a manufacturing environment where 100’s parts are required to be welded at the same time in low batch runs, welding only two at a time can’t meet the utilization criteria. More time is spent on part change over than actually producing. No arc, no revenue.

      When the tool has the capability to be brought to the skilled welder, moved throughout the shop where it is needed most, it becomes a force multiplier of the skill and production capability of the welder, not a system that needs to be fed. When the tool is freed from the constraints of one location in the shop and the tool can work in any booth, besides the welder, it is a tool and not a fixed system. The shop's production doesn’t change, it is made more productive. The tool is not the axis around which the production orbits, like traditional automation, it simply enhances every aspect of the way you produce products today.

      Short Delivery Times

      When the tool can be implemented in the same time frame as the welder needs to prepare for production, the tool has the potential to increase the throughput of the individual welder with no impact on the delivery timeline.

      Typical high volume/low mix production has long-term contracts stretching out over the years. This affords the luxury of time to develop the production strategy, specify and purchase the equipment, install and execute a startup plan. Low volume/high mix production usually has a short time between receiving the order and delivery. There is no time to specify and purchase equipment nor develop and build tooling. Manufacturing must start upon receipt of order. Traditional automation falls short. Even if the automation cell is onsite, tooling the parts and creating the program is usually longer than the timeline allows.

      When the tool can be implemented in the same time frame as the welder needs to prepare for production, the tool has the potential to increase the throughput of the individual welder with no impact on the delivery timeline. If the tool and welder are both producing the same part at the same time, the potential for increased throughput with no additional floor space, staff, or overhead costs opens the possibility for adding additional work with little to no additional expense to the company.

      What tool makes this possible?

      This tool has been made possible by Universal Robot’s non-traditional, industrial robot known as a collaborative robot, that is used in solutions like the one that will be introduced by Hirebotics in the near future. 

      A collaborative robot is one that can work side by side with people, interacting with them in various ways to accomplish a given task. When combined with a welding machine, welding torch, modular tooling system, and table/method of affixing to a part, it becomes the tool we have been discussing. This tool is free to move about your manufacturing floor, going where it is needed. The ability to “lead” the robot through its path makes complex programming and long training sessions obsolete.

      The Hirebotics' cloud platform has simplified robotic welding by unlocking unique capabilities for welding, delivering application-specific skills to its robots, and enabling revolutionary experiences for everyday end users through the use of the Hirebotics' mobile app.


      Traditional robotic welding has been the savior for many manufacturers. The benefits I discussed are real and they have experienced them. To maximize the potential of automation these companies have built automation departments to honor it and automation is the axis on which their companies turn. This is good, it is the reason cars are better today than ever before, a single farmer can buy and own a tractor for a lifetime and we still manufacture goods in America today. But it is an exclusive club.

      Download the Ebook

      Mitch Dupon
      Post by Mitch Dupon
      August 20, 2019
      Mitch Dupon is an Application Development Engineer for Universal Robots developing the collaborative welding market. Prior to joining Universal Robots, he spent 10 years with KUKA Robotics in their welding robotics group. He started his career with Panasonic Factory Automation after graduating from Welding Engineering at Conestoga College in Canada 1999. His experience includes installing, servicing and selling welding automation, from custom lines in high production automation environments like automotive to simple pre-engineered cells at small and medium sized businesses.
      Download the Ebook