② Smart Factory Korea: Examine the System Before Adding Headcount | Productivity Revolution for Business Leaders
Updated: Sep 27
Season 1 · The Beginning of the Productivity Revolution
By HOSOON CHOI
“Two more operators should solve the delay.” That may be the right decision. But first, identify what is preventing the existing team from completing more good work. Are people fully occupied at a constrained process, or are they waiting for materials, instructions, equipment recovery or quality release?
Headcount affects capacity. The operating system determines how that capacity is used. For a manufacturer establishing or expanding operations in Korea, separating those questions helps avoid recruiting for a problem that also needs a different solution.
Output and labor productivity answer different questions
Output measures how much you produce. Labor productivity measures that output against labor input. Neither, by itself, tells you whether an investment is profitable.
Consider an illustrative shift producing the same item to the same quality standard. Each operator works eight hours, and both scenarios use the same labor-hour accounting boundary.
Measure | Existing team | After adding two operators |
Operators | 10 | 12 |
Total labor hours | 80 | 96 |
Good units produced | 800 | 850 |
Good units per labor hour | 10.00 | 8.85 |
Output rises by 6.25%, while labor productivity falls by approximately 11.5%:
Labor productivity = good units produced ÷ total labor hours
The extra 50 units might still be commercially worthwhile—for example, if they meet a valuable delivery commitment. The calculation does not establish otherwise. It tells management to investigate why a 20% increase in labor produced a smaller output increase, then compare contribution, labor cost, quality, delivery and other relevant costs.
For mixed products, unit counts can mislead. Compare a consistent product mix or use documented standard labor hours earned, with the same standards and scope in each period. Do not make performance appear better by excluding support work from only one scenario.
Find the constraint before accelerating the line
Imagine three sequential processes with capacities of 60, 50 and 40 good units per hour. In a simplified steady-state model—with sufficient input, no additional downtime, losses or variability, and no alternate route—the line cannot sustain more than 40 units per hour.
Increasing the first process to 70 does not increase that ceiling. If upstream production continues faster than downstream consumption, work-in-process inventory accumulates until storage or operating controls limit it. Raising the third process from 40 to 50 could move the constraint to the second process, giving a new theoretical ceiling of 50.
A real line needs more evidence. Examine processing time, changeovers, failures, material starvation, blocked output and quality holds. Buffers, scheduling and product mix also affect actual throughput. The process with the longest nominal cycle time is a starting hypothesis, not an automatic diagnosis.
Five operating systems worth examining
1. Standard work
Agree the sequence, required materials, quality checks, safe method and response to exceptions. Develop it with the people doing the work, and record where product variants require different instructions. A standard should support training and consistent execution without pressuring operators to skip checks to meet an unrealistic time.
2. Constraint management
Record what stops the constrained process from completing good output. Address repeated losses there first when throughput is the objective. Improving another process can still matter for safety, quality or cost; be clear about the benefit you expect instead of assuming every local speed increase raises factory output.
3. Material replenishment
Define the trigger, delivery quantity, container, destination, responsible person and exception rule. Check whether the receiving location can accept the delivery. A reliable manual replenishment route may be appropriate. Automated guided vehicles (AGVs) or autonomous mobile robots (AMRs) become candidates when the task and operating conditions support them.
4. Information flow
Identify the authoritative work order, inventory status and quality-release record. Decide who updates each record and who acts when it changes. A manufacturing execution system (MES), enterprise resource planning system (ERP) or warehouse management system (WMS) helps only when responsibilities, status definitions and interfaces are clear.
5. Abnormality response
Separate equipment failure, missing material, quality hold and instruction delay rather than recording everything as “other.” Assign an escalation route and a recovery owner, then verify whether the action prevents recurrence. A dashboard is useful when its information changes a decision or prompts timely action.
Measure the work, including the exceptions
Start with five groups of observations:
Good output, rejects and rework by product and shift.
Labor hours, processing time and waiting time by process.
Work-in-process quantity and how long it remains between processes.
Frequency and duration of material shortages, stoppages and quality holds.
Operator and vehicle trips, their purpose, distance and completion time.
A short observation period can identify questions. Use representative shifts, product mixes, peak periods and abnormal conditions before sizing equipment or approving investment. Distinguish an event's elapsed time from the labor hours affected, and avoid counting the same wait twice.
Necessary inspection, traceability and safety activities should remain in the process. Look for redundant entries, avoidable searching and unnecessary repeat handling around them. Reduced recorded time is not automatically a cash saving; explain how the released capacity would actually be used.
Turn the evidence into an improvement decision
Match the intervention to the cause. Missing location information may call for location discipline and barcode confirmation. Late replenishment may require a clearer trigger and route. A genuinely labor-constrained operation may need recruitment, training or reassignment. Stable, repeated transport may justify an automation study.
Test a change on a defined scope and compare good output, labor input, delivery performance and quality over comparable periods. Include transition time and support effort. For automation, review pedestrian and vehicle interaction, load stability, stopping and emergency response, maintenance access and manual recovery. Site-specific risk assessment and validation remain necessary before operation; faster movement alone is not an acceptance criterion.
Agree the Korea project brief across all partners
An overseas headquarters, Korean operating team and local integrator need one shared brief before requesting equipment proposals.
Scope and site data: define the process boundaries, products, load units, demand profile and target measures. Share current layouts, observed cycle times, routes, access limits and operating windows. Identify measurements still requiring a site visit.
Interfaces and roles: document handover points between production, storage and transport, together with equipment signals, data ownership and system access. Name who approves global standards, verifies Korean site conditions, resolves exceptions, trains operators and accepts the completed work. Agree working languages and document versions.
Procurement constraints: disclose the budget range, required quotation currency, delivery schedule, shutdown windows, utility and network responsibilities, spares and local service expectations. Assign responsibility for imported-equipment logistics and any project-specific compliance review. Record exclusions and assumptions so bids can be compared on the same basis.
Detailed engineering, a request for quotation (RFQ), return-on-investment (ROI) analysis and acceptance testing should follow from that agreed information. A supplier's indicative proposal cannot substitute for missing site measurements or an agreed operating requirement.
Start with the cause of the lost output
Before adding people or equipment, establish where good output is constrained and why. Then compare staffing, process changes, information improvements and automation against the same operational objective. The aim is to help people complete useful work with fewer avoidable interruptions.
For an initial discussion, prepare your product mix, shift-level output and labor hours, suspected constraint, replenishment method and principal site restrictions. Further engineering, supplier coordination, RFQ preparation and ROI work can be scoped around the available evidence.
Request a free quote and consultation through the platform in Korean or English, or contact hci@hosoonchoi.com.
Comments