Fighting Flaws

Robots provide the precision needed to minimize part defects, improving quality


In today’s fast-paced and competitive manufacturing landscape, quality control is necessary to ensure products meet the highest customer and safety standards. As a result, manufacturers are increasingly using advanced technologies coupled with high-performance robots on production lines, achieving consistent results that meet exact specifications.

Extremely beneficial to fabrication processes, the repeatability factor that robots bring to the welding table – or any application for that matter – is perhaps the most intriguing advantage to robot users looking to keep quality in check. Programmable to apply the same parameters, such as torch angle and travel speed consistently every time, extremely efficient robotic arms and their peripherals work together to keep part inconsistencies caused by worker fatigue, piece-part deviation and more at bay.

The use of robots and advanced vision and sensor capabilities for part testing and inspection maximizes good parts, reduces scrap and optimizes overall equipment effectiveness.

Ideal part types

Contributing to key performance indicators such as reduced waste and lower costs, robots provide the precision needed to level-up product quality and minimize defects for a variety of part types, such as:

  • Safety critical. A term widely used for the manufacture of automotive and aerospace components, safety critical parts (e.g., suspension systems, body structures and seat assemblies) are designed to withstand specific structural and safety requirements that are typically designated by a welding procedure specification (WPS).
  • Structural. Essential components for things like buildings and machinery, structural parts are often considered critical to the greater assembly of a given fabrication or even a building. Whether a beam, bracket, coupler, fastener, gear or something else, if welds are missing from any of these vital parts, the integrity of an entire structure is in question.
  • Aesthetic. Whether the final product is a yard ornament, playground decoration or metal toy, customers demand visually appealing welds. Jagged seams or welds with a copious amount of spatter provide a poor visual representation, and at times, can even pose safety risks.

Regardless of the part type, any engineered part should hold to a WPS, as mentioned. A WPS is a formal document that describes welding procedures for the creation of consistently identical welds that are designed for a specific part. These specifications are developed for each material alloy and welding process used, and it is generally driven by a preferred welding technique or specific code used at the production facility. Any part not meeting WPS criteria should be rejected, and measures should be taken to correct any discrepancies through scrap or repair of parts.

QC inspection

The typical inspection of many parts is achieved by taking a sample – such as, one out of every 20 or 200, for example – then sending it to a human worker or workstation, where instruments and criteria are used to deem parts acceptable. The human approach, however, opens the door to subjectivity. Certain weld bead characteristics, for example, might pass one individual’s standards but fail another’s.

Companies like IBM and Yaskawa are teaming up to enable a high-resolution thermal camera to capture data on the live weld pool, while the sound of the weld is also recorded. Photo courtesy of IBM.

In contrast, robotic automation enables the potential for 100 percent inspection of parts, ensuring that every item meets the required pass/fail standards. Robots equipped with sensors can perform faster and more thorough inspections than humans, identifying defects with greater accuracy to mitigate recalls and field failures. At the same time, serialization and data collection can now happen for each part, improving weld traceability.

This data collection allows for investigations to find any potential fault causes later in the product life, then pinpoint possibly affected parts. In automotive, for example, this is the difference between recalling dozens of vehicles or thousands of vehicles.

Sensing data

Sensor technology paired with intuitive software and a user-friendly interface can drastically improve consistency and overall part quality. For robotic welding, sensors typically fall into four categories: touch, through-arc, laser and vision. Likewise, they have three primary functions: seam finding, seam tracking and part scanning – all of which can be used for the process of inspection. Each function offers unique benefits, depending on the part and expected outcome, and most of the technologies used can be mixed and matched, where use is not redundant.

Accessible from the robot teach pendant, weld interfaces often have documentation and data analysis systems capable of connecting multiple power sources for comprehensive parameter evaluation. Hundreds of times a second, a power source measures its output voltage, wire feed speed, weld time and other determined parameters, comparing collected data to predefined limits for the weld in progress. If a power source discovers a parameter outside of acceptable limits, a weld can be marked as, “suspect,” and it can be handled properly.

Many applications use Miller Insight Centerpoint arc data monitoring software in conjunction with a powerful Auto-Continuum power source for this. Similarly, Lincoln Electric offers Production Monitoring with a unique WeldScore feature, and Fronius provides a tool known as WeldCube that can connect multiple TPSi welders from one device and can generate daily quality reports for equipment status.

In a time when customers demand visually pleasing welds, welding with high-performance robots helps provide improved aesthetics, along with additional benefits.

Data can also be generated from external devices. Laser profile sensors are ideal for quality inspection, and high-performance cameras can be used for reading serial or QR codes – with all data being compared to predetermined thresholds to determine acceptable versus flawed parts. Acoustic and infrared cameras are also being used more frequently during the welding process to detect errors before the robot is even completed welding.

Next up

Moving forward, advances in artificial intelligence, modular hardware and simulation will continue to enhance robotic applications, positively impacting real-world challenges. Aside from verifying part quality for a wide range of products, robotic inspection can also be used for a variety of tasks in hazardous environments such as in a foundry, underwater or where high radiation is present to protect human workers from unneeded risk and exposure.

Overall, end users and customers want to know that the products and structures they are using are safe and reliable. With plenty of tools available today, there is no reason not to invest in technology that protects workers and customers and avoids unnecessary litigation due to premature product failures or even injuries. Through more repeatable, reliable and traceable processes, robotic automation offers manufacturers the means to provide the assurance required. In turn, this usage enables greater scalability, impacting production in more ways than one.

Yaskawa America Inc.

Get industry news first
Subscribe to our magazines
Your favorite
under one roof