How to Kill a Robotic Project in Five Easy Steps

July 12, 2001
By: Jim Berge

How would you like to fail utterly in your quest to modernize your shop with robotic weldingequipment? Look no further—we've got all the bad advice you need right here.

The same few roadblocks always seem to conspire to either cause a robot project to fail or at least greatly limit its success.

Too many manufacturers do things that virtually guarantee that their robots will have serious problems performing as expected. Many of these hindrances have simple solutions, but these solutions often are ignored because they require a little bit of effort or money to implement. It's almost as if some manufacturers want their robots to fail.

To that end, here are some of the best things-drawn from real-life applications-that you can do to ensure the failure of your robotic welding project. Follow these five steps religiously, and your robotic system will never meet your expectations.

Step 1: Don't Involve the Shop Floor Workers

If anyone in the factory knows the process intimately, it's the person on the floor who welds the products day in and day out. Workers are the ones with the process and product knowledge, the experience to know what works and what doesn't, and the opinions about how to improve the process. Deny these people any participation in the robotic welding project, and it surely will be less successful than desired.

If workers are not involved in the configuration and specifications of the equipment up front, chances are that the equipment will not be as ergonomically friendly or as productive as it could be. If they are not involved in the decisions about process and fixturing, then the little tricks and insights they have learned through the sweat of their brows will be lost.

If workers are not trained on how to program and run the robots, their welding and process knowledge will not be transferred to the robot's programs. This, too, will help to cause the robot system's failure.

Step 2: Pay No Special Attention to Who Your Robot Operators and Programmers Are

The robot industry follows the computer industry advances closely. Even though many robot manufacturers concentrate on developing robot languages that are simple, intuitive, and easy to learn, end users are demanding more flexibility and power from robot languages and controls.

A computer-literate individual can learn a robot's language quickly and instruct the robot exactly how it should move and act to accomplish the desired welding tasks.

Self-motivated employees who want to learn and advance their skills and knowledge are the ones who should become robot operators, programmers, and technicians. Of course, programmers also should be good welders, because their primary collective focus is to transform their welding knowledge into actions performed by the robot. Choosing a robot programmer based solely on, say, seniority, with no consideration of the individual's attitude or skills, is an effective method of short-circuiting a robotic welding project.

Step 3: Make Training a Low Priority

As important as choosing the correct person to program your robots is, choosing the type and extent of training that person receives is just as critical. An effective robot programmer possesses a logical mind that is able to understand cause and effect quickly. This person has an intuitive sense of how to make things work, how to solve problems, how to improve existing processes, and how to create something out of nothing. With the proper training, this person is well equipped with the tools to create fast, efficient, and effective robot programs.

A multilevel training regimen is required to bring a programmer to this level. Basic programming training provides only groundwork on which subsequent training builds. The real learning happens on the shop floor and during advanced training courses.

In addition, advanced training is necessary for a programmer to fully utilize a robot's abilities. This may include learning how to make the robot communicate with other equipment; how to make the most use of certain options-for instance, touch sensing or seam tracking; how to optimize cycle times; or how to use the robot for tasks other than welding parts (e.g., counting number of cycles and tracking cycle times).

All of these things can add to the productivity of a robotic welding cell. So, it stands to reason that if your goal is to kill your robotic welding project, then simply ignore training altogether. Almost as effective, however, is to provide programmers and technicians a basic training course and nothing else.

Step 4: Pay No Attention to Part Fit-up and Repeatability

The single most common problem confronting welding robots is part fit-up and repeatability. It is clear that neglecting these crucial factors negatively impacts the reliability of the process and causes significant quality problems.

Robots can use touch sensing or seam tracking software to compensate for weld joints that move around. But the ability to deal with gaps is limited by the welding process itself, not by the robot. Simple physics declares that molten metal will not stick to air, which is all that exists in a gap between parts.

Many companies that have experienced success with robotic welding systems have included the fabrication of parts in the same capital project as the robotic welding system. Along with the welding robots, they also purchased lasers, plasma torches, and computerized press brakes with which to build high-quality, repeatable components. The companies that excel at implementing welding robots understand the value of presenting repeatable parts to the robots. They understand that presenting inconsistent parts to the robots that have big gaps between components can bring them one step closer to failure.

Step 5: Assume That Throwing Extra Labor Into Your Welding Department Is Just As Effective As Installing Robots

Figures from the Robot Industries Association (RIA) show steady growth in the robot industry for the past 10 years and project steady future growth.

Because welding comprises a significant proportion of all robot applications, it is safe to assume that your competitors are buying welding robots. The average welding robot can do the work of three to four people, so you can be assured that costs will increase significantly if the choice is made simply to throw more people into the process.

If an increase in orders or a new product introduction is pending, it makes a lot of sense from a financial point of view to consider welding automation. It also makes a lot of sense regarding efficiency, productivity, and quality. Higher quality will create higher demand for your products. Better efficiency makes your product more competitive. And higher productivity allows you to meet production demands without adding lots of people to the process.

Skilled welders are becoming very difficult - even impossible, in some parts of the country - to find. By depending more and more on manual labor in your welding department, you inch one step closer to failure.

Follow these steps carefully, and you can rest assured that whatever other efforts you take to install welding automation and improve your welding operations will all be in vain.

Jim Berge

Jim Berge

Contributing Writer
Berge Robotics
Sherveport, LA
Phone: 970-222-1753