Measuring Return on Automation

When operations managers and engineers evaluate how to improve manufacturing processes, they often look at implementing new or improved automation systems.

After careful studies of the costs of purchasing hardware and software, and adding in the costs of engineering and construction, they are ready to make a request for funds from the business side of the company. It is very likely that they will hear the question, “What is the expected return on this investment?” In fact, if they are unlucky, they may also hear financial management ask questions such as, “As a matter of fact, while we’re on this subject, just what has been the return on all the automation that we’ve bought over the last five years?”

Engineers are often taken aback by these questions. They’ve studied the technology behind the systems and how to engineer the implementation. They know the operational benefits the system is designed to bring—improved quality, less waste, reduced manpower, increased productivity, faster product changeover and the like. They may not have even thought about the larger financial ramifications to the corporation.

Peter Martin, vice president of performance management at automation supplier Invensys Process Systems, Foxboro, Mass., has been studying this problem for well over a decade. He published some of his discoveries in “Bottom-Line Manufacturing,” published by the Instrumentation, Systems and Automation Society (ISA), now in its second edition. Martin’s thesis on determining returns on automation begins with considering the lifecycle.

“To calculate returns, you need more than lifecycle costs,” he says. “You need to have an accurate measure of the lifecycle benefits the manufacturing operation derives from using the automation system.” Martin’s extensive interviews with manufacturing executives revealed that none actually measured benefits created by automation. In fact, “most admitted that they did not know how to get at these metrics in any reasonable way, and that their current cost accounting systems did not provide detailed enough economic data to be able to infer the benefit value.” Economic benefits typically accrue either from reducing manufacturing costs, increasing production, or a combination. Martin found that the only metric typically followed was reduced manpower requirements, because it was the most easily measured.

Measure to control

Perhaps the basic tenet of process control can be stated as, “If you can’t measure the controlled variable, then you can’t control it.” Companies have not done a good job of measuring benefits that accrue from automation implementations. Martin discovered three barriers to effective economic lifecycle benefit achievement of automation in manufacturing.

He calls the first one the “replacement automation” approach. This occurs when engineers begin to evaluate replacing obsolete automation or process control equipment. Typically, they will build a requirements or specifications document that is constructed from the system they already have in place. Unfortunately, this specification ignores all of the development work done by suppliers in the industry to bring new technologies and efficiencies into the market. The procurement team, instead of looking at performance enhancements, focuses on replacing what they have with the same technology at a lower price.

The second barrier is what Martin calls the “project team approach.” In this analysis, the company establishes a team of engineers and other personnel to evaluate, specify, procure and implement the new automation. This group acquires considerable expertise in the system, its capabilities, its potentials. But, when the project is successfully implemented, the team is disbanded. The company loses all of that institutional knowledge and the eventual benefits are never realized.

On the other hand, the company never realizes it has lost that benefit due to the third barrier—it never measures the benefits.
Return on automation = lifecycle benefits – lifecycle costs.
To know the return on automation, you must measure both.

Gary Mintchell, gmintchell@automationworld.com, is Editor in Chief of Automation World.

 

 

More in Control