Two denominators
Productivity is output divided by input. Which input you choose changes the story. The two most common choices in plants are labor hours and machine hours.
- Output per labor hour asks how much each hour of paid work produces.
- Output per machine hour asks how much each hour of equipment time produces.
Automation usually raises one of them more than the other, and sometimes pushes them in opposite directions. Seeing both avoids drawing the wrong conclusion.
A hypothetical calculation
The numbers here are invented to show the arithmetic and do not describe any real plant.
Before automation, a packing line produces 120 units per hour and is staffed by 4 workers.
- Output per labor hour: 120 divided by 4 is 30 units per labor hour.
- Output per machine hour: 120 units per hour of line running time.
After a project that adds automatic case loading and a label checker, the line makes 150 units per hour with 3 workers.
- Output per labor hour: 150 divided by 3 is 50 units per labor hour.
- Output per machine hour: 150 units per hour.
Labor productivity rose from 30 to 50, an increase of about two-thirds, while machine productivity rose from 120 to 150, an increase of 25 percent. The labor figure moved more because the project reduced staffing as well as raising speed.
Now consider a second possibility. Suppose a different project adds a very capable but expensive machine that runs at 150 units per hour but requires 4 workers to supply and monitor it, plus a technician for part of the day. Output per machine hour improves, but output per labor hour is 150 divided by 4, or 37.5, and if the technician’s time is counted, lower still. The machine hour figure alone would flatter the project. For a full picture, you might also consider output per unit of cost, but that requires real cost data, which is beyond this example.
What automation changes
Speed and rate. Machines can run faster or more steadily, which raises output per machine hour.
Machine time used. Automation that loads parts, monitors itself or restarts after minor stoppages can raise the share of scheduled time during which the machine produces. Equipment uptime often improves more from tackling small stops than from raising top speed.
Labor per unit. If tasks such as loading, moving and inspecting are handled automatically, fewer labor hours are needed per unit, or the same people can oversee more machines.
Quality. Output per hour should count good units only. A line that makes more units but scraps more has not become more productive. Automated quality control matters here.
How the work of people changes
When a repetitive task is automated, the people who did it do not necessarily disappear from the process; their work changes. Typical shifts include:
- from loading and unloading to supervising several machines
- from manual checking to responding to inspection results and investigating causes
- from fixing breakdowns to carrying out planned maintenance
- from following a routine to helping with setup, programming and process improvement
These roles generally require more training. They may also come with more responsibility, and not every person will want or be able to make the move. Employers should plan for training and be honest about effects on staffing. The wider labor-market question is discussed in automation and jobs; individual outcomes vary by company, region and industry.
The phrase “higher-value tasks” should be treated carefully. It is a hoped-for outcome, not an automatic one. If the plant simply cuts hours and never uses the freed capacity of staff for improvement work, the value gain is limited to labor savings.
Pitfalls in measuring
- Counting the wrong units. Always separate good output from total output.
- Ignoring support labor. Maintenance, programming and supervisory time that supports the automated equipment is part of the labor input.
- Mixing time periods. Compare similar product mixes and operating conditions; a change in what is made can look like a change in productivity.
- Short test periods. Early results after start-up may not reflect steady operation.
- Hidden constraints. Output per hour at one station may not matter if the constraint is elsewhere.
Practical advice
Decide in advance which measure matters for the decision, record a baseline over a representative period and measure again after the equipment has settled in. Present labor and machine figures side by side so that tradeoffs are visible. If you want ongoing improvement, follow-up measures after the project are discussed in the article on productivity improvement.
This article is general information, not engineering or financial advice.