Whether you’re an automotive manufacturer, retailer, or consumer, there’s one word you don’t want to hear when talking about quality: recall. In 2022 alone, over 1,000 separate recalls were issued worldwide, affecting over 30 million vehicles. The cost is unquantifiable—and the damage to a brand’s good name in the case of high profile recall is unimaginable.
While recall is common, preventing it remains a top priority at automotive manufacturers around the world. Teams of skilled staff—from QA and maintenance managers to robot programmers and beyond—all work together with cutting edge tools and processes to combat quality issues.
So why does recall remain a seemingly unsolvable challenge? In order to get to the bottom of the issue, it’s worth a closer look at how recall-worthy defects occur and why.
According to the U.S. National Highway Traffic and Safety Administration, among the most common causes of recall are issues related to “improper processing,” “mis-located processing,” and “missed processing”. These kinds of recalls occur when there is a problem with (typically automated) joining operations: a missing weld or glue seam, for example. In the simplest scenario, a car rolls off the line missing a weld, the issue isn’t caught, and the vehicle goes to a consumer—where disaster can occur.
So just where in the manufacturing process could a missed processing recall have been prevented? The problem typically has its roots in the very beginning of the manufacturing process: robot programming. In large manufacturing plants, it’s common for changes to be made to robot programming: staff and subcontractors constantly adjust for optimization and other reasons. These ‘quick changes’ are often undocumented—and it’s easy for an error like swapped IDs to slip in with routine cycle time optimization. When you come into the factory on Monday morning, there is often no record of what changed over the weekend. Production starts with no way for all staff to understand the current plant status—until a bad part is produced and detected.
Issues like missing welds are often detected through destructive testing: the process of physically dismantling a vehicle to examine the quality and integrity of the assembly process. An issue like missing or improper welds would be immediately apparent to maintenance professionals. But less easy to see is where and how to find the root problem. Documentation of where, when, and how each process is done is often difficult to access—doubly so for more specific process data like weld parameters. So even once an issue is detected, implementing a successful fix can be complicated.
That’s only a quick look at some of the common challenges posed by robotic welding, and it’s not all: faulty commissioning, complex variants, and staffing challenges also contribute to the problem. The complexity and scale of data generated by robotic manufacturing is a challenge to maintaining consistent quality from automated processes. But it doesn’t have to be: RoboLive® is a software solution designed to combat these challenges and allow you to take better control of your industrial robots.
RoboLive® is installed in the production network and runs automatously, collecting robot data each shift. With an installable client combining process visualization, data aggregation and analysis, it’s an easy way to keep track of production status instantly.
Automatic change tracking is a big part of what RoboLive® offers. RoboLive® keeps the entire history of your production and makes it viewable and comparable on a timeline. At the start of a new week, day, or shift, a single click will show all programming changes to all staff. RoboLive® even tracks this data over time, allowing for automatic reports detailing which processes or robots receive the most frequent adjustments and when.
Automatic errors and warnings form a powerful first-line defense against quality problems. RoboLive® compares nominal from planning with actual data from the robots themselves each shift, and provides automated notifications when a process is out of line. Missing welds would warrant a critical warning, making the issue immediately obvious and easily solvable. And it’s not just missing welds: Off-location processing, unplanned processing, and even typically undetectable issues like swapped IDs are all detectible with RoboLive®, including custom errors to suit your manufacturing standards.
By establishing an automatic link between process, ID, robot, and physical part, RoboLive® also makes it easy to direct maintenance action and ensure quality results. Every process on your lines is represented in RoboLive® exactly in where it occurs in 3D space—and the results can even be algorithmically adjusted for bad base frames. The information is viewable to everyone in the plant at all times, via line PCs or office computers.
Both total recalls and units recalled only continue to increase each year, and it’s something that every manufacturing processional should be concerned with. As production lines rely increasingly on robotic processing and increasing amounts of data are generated, it’s more important than ever to have the tools to deal with it. RoboLive® is running on some of the largest automotive manufacturers—and no line it’s installed on has ever reported a recall.
Right now, our team is offering customized trial installations of the RoboLive® client, letting you view your own product data in the software solution, complete with an analysis meeting outlining any detected problems. After a brief exchange of data and a short preparation period, you’ll have a clear idea of what RoboLive® can transform your manufacturing—and how you can better arm your team against recall-worthy processing defects. Contact our team today to set up your consultation, and follow us on LinkedIn for latest updates.