When Hours Shrink, the Production Plan Breaks First
As of the end of June 2026, 224 companies had joined Korea's Ministry of Employment and Labor "Work-Life Balance +4.5" project, already exceeding the annual target. What stands out is that 67.9% of participants have fewer than 50 employees. Working-hour reduction is no longer a large-corporate HR topic — it is the reality for small and mid-sized manufacturers that generate revenue from a single line.
The friction appears where the scheme meets a continuously running line. Adopting a fixed 36-hour week with different scheduled hours by weekday — 7 hours Monday and Wednesday, 8 hours Tuesday and Thursday, 6 hours Friday — leaves equipment running as before while staffed hours vary day by day. Who covers the two-hour Friday afternoon gap, and how that capacity is reflected in the plan, becomes an immediate issue.
In plants that manage attendance in spreadsheets, every new work pattern adds both formulas and exceptions. Add weekday-specific hours to three work patterns, then mix in substitutions and overtime support, and company-wide payroll ends up depending on the one person who maintains the sheet.
What It Means to Define Shift Patterns as Data
Drawing a 2-shift, 3-shift, or 4-crew-3-shift schedule on paper is not the same as defining it in a system. A system needs at least four things made explicit.
When scheduled hours differ by weekday, the overtime threshold moves with the weekday. Working 8 hours on a Wednesday with 7 scheduled hours does not exceed the statutory 8-hour daily limit, so it is not statutory overtime — but if a collective agreement treats hours beyond scheduled time as premium-eligible, that portion needs its own line item. Hard-coding these thresholds as constants means a redeployment every time the rules change. Weekday baselines belong in the work-pattern master as data.
Crew composition changes, substitutions, and overtime support should be handled as exception records layered on the pattern, not as breaks in it. Preserving the original pattern while recording who covered for which crew under whose approval is what makes later settlement and retroactive recalculation possible.
Attendance Data Is Labor-Risk Data
Clock-in records are not just logs. Break-time deduction rules, paid-versus-unpaid classification, and how late arrivals and early departures are handled all feed directly into payroll. When reduced hours are compensated as paid time, the supporting record matters even more. If "reduction compensation" is not separated as its own pay item, interpretations will diverge months later over whether it belongs in ordinary wage calculations.
The disputed items are fairly predictable.
Given the three-year statute of limitations on wage claims, source attendance records and approval trails should be retained for at least three years, change history included. Storing the approver, the approval timestamp, and the pre-change value turns the data into something explainable during a dispute.
What Becomes Visible When Attendance Meets MES
"Time on site" in an attendance system and "labor hours" in MES are not the same number. Forcing them to match costs both systems their credibility. Designing indicators that explain the gap works better.
It varies by industry and process, but many plants treat a direct labor ratio in the 80–90% range as manageable. If it drops into the 60s, the signal points to line operations rather than attendance aggregation.
Headcount-based capacity should be calculated daily in sync with the work pattern. Moving from 40 to 36 hours reduces available hours per person by 10%, but by weekday the 8-hour Tuesday and Thursday are unchanged, Monday and Wednesday drop 12.5%, and the 6-hour Friday drops 25%. Averaging capacity across the week hides Friday delivery risk at the planning stage.
Three indicators show what absorbed the lost time: units per hour (UPH), downtime, and changeover waiting time. If four hours per week disappear and UPH stays flat, output genuinely falls. If changeover waiting ran two hours a week, cutting just half of it recovers a quarter of the reduction.
Sites Mixing Shifts, Contractors, and Dispatched Workers
When workers from different employers share a line, the system must separate three things: authorization (what prime-contractor staff versus partner staff can view), settlement (who calculates direct wages versus contract and dispatch fees), and safety training records (completion history and validity periods by employer). Merging these into one table blurs the basis for both settlement and safety compliance at once.
Input hardware follows site conditions.
Most failed rollouts start with exceptions that go unrecorded. When a forgotten tag, a terminal outage window, or a hastily assigned helper ends up on paper or in a chat message outside the system, the plant drifts back to running spreadsheets in parallel. Design the exception-entry path first.
A Phased Transition Roadmap
Changing everything at once is the riskier path.
Each step should begin only after the previous step's data quality is established. Step 3 in particular warrants at least two months of parallel operation against the existing method.
Drawing on our MES and internal-systems experience, POLYGLOTSOFT designs these systems on the assumption that shift patterns and attendance rules will keep changing. Instead of redeveloping every time a regulatory change or collective agreement adds a work pattern, our subscription development model keeps a dedicated team continuously reflecting new rules and extending into MES integration. If you are redesigning attendance and labor-hour systems ahead of an hours reduction, we will review a phased approach with you based on the work patterns and line conditions you run today.