Automation is spread across the entire factory, from logistics to production. Six-axis robots are used for welding and screwdriving. Automated guided vehicles transport parts and finished appliances throughout the plant. And, all of the plant’s manufacturing processes are overseen by LG’s own digital-twin monitoring system. The system uses artificial intelligence and “big data” to assess inventory status, production performance and machinery failures.

Because automation has freed employees from dirty, dull and dangerous jobs, the plant has not struggled to fill job openings.

Less well-studied is how automation might affect the skills of factory workers. Now, a new study is shining some light on that issue.

Matthias Rötting, Ph.D., a professor of industrial engineering at the Technical University Institute of Psychology and Ergonomics in Berlin, and Zahid Rashid, Ph.D., a professor of industrial engineering at the University of Engineering and Technology in Taxila, Pakistan, examined the effect of automation on the sewing skills of workers at a garment factory. The researchers hypothesized that workers who spent extended periods tending automated sewing machines would gradually lose their ability to operate manual machines over time.

Sixty-seven garment workers were trained on manual sewing machines to perform a complex production operation. Then, the workers were divided randomly into three groups exposed to varying amounts of automation. The manual group continued working on the manual sewing machines. In contrast, the automation group shifted to automated sewing machines. Workers in this group simply had to load and unload automated machines, which did the sewing work for them. The third group, the refresher training group, rotated between manual and automated machines.

The researchers assessed the manual sewing skills of all the workers once after their initial training and again after six weeks at their production assignments.

The researchers found no significant differences in performance between the three groups after their initial training.

After six weeks, however, performance differences emerged between the groups. In the automation group, average single cycle time increased, while the right-first-time percentage decreased. In the manual group, the results were just the opposite: Average single cycle time decreased, and the right-first-time percentage increased. That makes sense, of course. After six straight weeks of practice, one would expect workers in the manual group to be better at manual sewing than those in the automation group, who hadn’t touched a manual machine in that time.

The refresher training group showed no differences in performance between initial training and six weeks of work.

The results should provide food for thought for plant managers and line supervisors. Anyone who has ever tried to learn a foreign language is familiar with the phrase, “use it or lose it.” If you don’t speak or read the language on a daily basis, you’ll likely forget what you’ve learned. The same could apply to manufacturing skills—what in another day and age might have been called craftsmanship.

There’s no doubt that robots, automation and error-proofing technology are great tools for boosting productivity. We must continue to invest in such technology to remain competitive. But, we should also be mindful of creating workplaces that are sterile and oversimplified. Employers are well-advised to institute regular job rotations and to continually invest in workforce training. Skilled workers who are engaged in their jobs and proud of their work will produce quality products, and that, too, will boost the bottom line.