Software is Cheaper Labor

Two remote calibrations running on a $100 compute stick, using Metrology.NET software.

Over the years I have watched several labs go through the decision process to purchase new hardware. Most labs have the same process: look at the work load requiring the new hardware, and calculate the payback of new equipment. Upper management usually wants to know what the Return on Investment (ROI) looks like. And the numbers usually speak for themselves. The only variable is usually the time frame—how soon does spending this money save the company money?

If getting a new calibrator for the lab would end up making the lab more money, it would be a good investment. So if a new calibrator costs $60,000, with an annual calibration of $4,000, it costs a total of $64,000. If the company wants to see a ROI within one year, the manager would have to show the $64,000 will make the company at least $64,000 within one year.

As a programmer, what has always annoyed me is how software has such little value compared to hardware! Management gets sticker shock for software but not for hardware.

So let’s continue with the above example. Buying a new calibrator can at best double the workload. Whereas software can double, triple even quadruple the workload. In fact, 4x the throughput with automation is not uncommon—it’s the average. So why is software so undervalued?

I recently quoted a customer $4,000 for a Yokagawa WT3000 calibration procedure. This was the first time I quoted this customer and I know it was sticker shock. But a manual calibration of this hardware is six hours, while automation can run in just one (6x the time savings). A customer supporting 10 of these units could save $6,250 per year as $125/hour loaded shop rate. The math is simple: 10 * 5 * $125 (#Units * SavedHours * ShopRate). The software would pay for itself after just seven calibrations.

I understand if a lab only supports one or two of these units; the justification is not there, but the argument for spending money on automation should be easier than purchasing new hardware. In the above example, you still have five hours you can use that hardware for other work. Whereas if you would have spent money on new hardware before software you would have the additional cost of man hours to operate the hardware.

I had this conversation with a lab manager I have known for years, and not surprisingly, he was in 100% agreement with me. He explained it from a lab management perspective: the real value is not only in the time savings but also in the cost of labor. Automation allows him to hire less skilled labor and that saves money in the long run—he is looking at the long-term problem the industry is facing.

His focus is on finding and training the next generation of calibration technicians, as he loses his more experienced technical staff. How will his lab operate 10 years from now? He envisions automation at the center. Because of the military draw down 20 years ago, there are less highly qualified calibration technicians. But with software, he can hire less skilled technicians, put them in front of a computer, then use his higher skilled technicians to problem solve and train the newer people—a logical division of labor, like the doctor and the nurse.

This lab manager doesn’t see the purchase of software as a capital expense, but rather as a labor expense. So, the $4,000 for the software gets wrapped into the $125/hour loaded shop rate. If he can pay a technician $17 per hour vs. $35 per hour, that gives him a budget of $37,500 per year to spend on software. I thought this was genius, I’ve always said “Your best employee was a $1,500 computer with a GPIB card,” but never took the time to do the math per technician costs as opposed to per procedure cost justifications.

Software is not only cheap labor, but will save you calibration time in the long run, reduce hardware costs, and lower labor costs of a high quality calibration lab.