Client
Bossard Singapore
Role
Operations intern
Discipline
Process analysis

Schedules were built on a capacity nobody had measured

The benefit

What was achieved

Cost savings
17%
Reduction in overtime post accurate planning and promising
Process improved · knowledge
100%
Of receiving and packing documented, from nothing
The problem

Packing capacity had never been measured, so schedules committed more work than the floor could do and overtime covered the gap.

My contribution

What I personally did on this

Owned
The time study · capacity analysis · both written procedures
What was delivered

A procedure, where there had not been one

What existed before
  • Nothing written. The procedure lived in whoever happened to be on shift
  • Taught by demonstration, so each trainer passed on a different version of the same job
  • No failure paths. A shortage, a failed weight check or an item pulled to quality meant asking someone
  • System entry learned by watching another person type. No record of which field takes what
  • No way to tell when a new hire was ready, so supervision continued by feel
What was delivered
  • Receiving flow, end to end, with four decision branches for quality routing, shortages, PO completion and outside service
  • Packing flow with the multi-unit scan loop and the weight-control gate
  • Five system steps for keying lot information, field by field, with the real screens
  • Ten numbered physical steps through to a sealed and documented box
  • Local and overseas shipment split written down, including where the delivery note goes
Receiving process flowchart with decision branches
Receiving — the four decision branches are the part that used to require asking someone
Packing process flowchart with scan loop and weight control
Packing — scan loop, weight-control gate, staging for forwarder collection
Workflow analysis

Where the bottleneck lived

Before
Shift planned on a 480-minute day
→
⚠ Bottleneck Nobody had ever measured a cycle no cycle time · no written procedure · the floor ran on memory
→
Dates slip, overtime absorbs the gap
After
Same operation, one stopwatch
→
✓ Element study Cycle split into find, pack, seal 179.4 seconds average across five cycles
→
✓ Real capacity 350 operating minutes, 2.05 days per pallet planning stops using a day that does not exist
Ongoing, without me
A new packer joins
→
✓ Written SOP Receiving and packing, end to end the operation stops depending on who is on shift
Notes

What shaped the decisions

The finding the whole project turns on

Twenty seconds of variation in one step costs sixteen minutes per pallet. That is the entire argument for standardisation, measured rather than asserted.

Why this was the deliverable that outlived the internship

A time study is a snapshot. A written procedure is the thing that stops the operation drifting back the week after anyone stops watching.

Two more things, neither of them asked for

An element-level time study that produced the cycle time and the capacity correction, and two floor improvements implemented — pallet boundary markings and speed-limit signage, both raised and closed the same month.

Requirements

What they needed, and why it mattered

Five requirements. Each one has a business consequence attached, and each consequence shows up in the numbers at the top of this page.

  1. 01
    Measure the real day, not the one on paper
    A shift is 480 minutes. After breaks and unavoidable non-productive time it is 350. Planning used the first number.
    Every schedule and every date given to a customer was 37% optimistic before anyone checked.
    Derived requirement · the blocker to everything else
  2. 02
    Break the cycle into elements, not one number
    A total time tells you the operation is slow. It does not tell you which part to fix.
    It named packing as 65% of every cycle, so improvement effort finally had somewhere to go.
    Derived requirement
  3. 03
    Prove whether variation actually costs anything
    "Everyone works slightly differently" is easy to wave away. Two runs of the same operation, one element changed, settles it.
    Twenty seconds in one step turned out to be 16 minutes per pallet. Standardisation stopped being a preference.
    Derived requirement
  4. 04
    Write the procedure so a new hire can run it
    Receiving and packing existed only as habit. There was no document to hand anyone.
    The operation was one absence away from slowing down, and training took as long as someone was willing to stand there explaining.
    Client need
  5. 05
    Fix what the measurement exposes on the floor
    Standing on the floor with a stopwatch surfaces things a desk never sees.
    Two safety hazards were found and closed the same month — undefined pallet boundaries and no speed control in a forklift zone.
    Derived requirement
Method

One stopwatch, five cycles

No new systems, no software, no disruption to the line. The whole method is a stopwatch, a structured sheet, and the discipline to run it twice.

Time study method
Pick the operation  →  split into elements  →  time 5 full cycles
                                    ↓
                      Average each element separately
                                    ↓
      cycle time  →  × 48 cycles  →  pallet time
                                    ↓
      Real operating minutes  →  cycles per day  →  days per pallet
                                    ↓
             Run it again, change one element only
01
Split the cycle. Repacking breaks into three elements — finding the item, packing, sealing. Time each one separately, not the whole.
02
Time five full cycles. Four small boxes plus the consolidation into the shipping box, so the sample covers the real mix of work.
03
Average per element. This is the step that turns a stopwatch into a diagnosis — a single total would have hidden where the time sat.
04
Scale honestly. Cycle × 48 gives pallet time. But cycles per day must use operating minutes, not shift minutes.
05
Run it twice, change one thing. A second study identical except the final seal isolates what variation actually costs.
Findings

Four gates, each one narrowing the search

Every step asks one question, answers it with the measured data, and only then moves on.

01How long does one cycle actually take?
The repacking line where the time study was conducted
The operation being timed — repacking, five full cycles
Time study — repacking, seconds per element
CycleFind itemPackingSealingTotal (s)
Small box 13510824167
Small box 22511228165
Small box 35011722189
Small box 44111832191
Consolidate into shipping box1513040185
Average33.2117.029.2179.4
What it showed

A full cycle averages 14.95 minutes. A pallet is 48 cycles. At the real operating rate that is 2.05 days per pallet — the first cycle time the operation had ever had.

02Where does the time actually go?
Share of every cycle, by element
ElementSecondsShare
Packing117.065%
Finding the item33.219%
Sealing29.216%
What it showed

Two thirds of every cycle is a single element. Any improvement not aimed at packing is aimed at 35% of the problem — which is exactly where an untargeted efficiency drive would have gone.

03Does operator variation actually cost anything?
Two studies, identical except the final seal
StudyFinal sealCycleCycles/dayDays per pallet
Run 140 s14.95 min23.412.050
Run 260 s15.28 min22.902.096
Difference20 s——16 min / pallet
What it showed

One element, changed by twenty seconds, moves the pallet by sixteen minutes. Nothing else in either study differed.

That converts "everyone works a bit differently" from a shrug into a number, and it is why standardising the seal step was the first recommendation rather than a nice-to-have.

04How much capacity did they think they had?
The planning day against the real day
Minutes in a shiftValue
Shift length480
Breaks — lunch, tea, toilet−90
Miscellaneous non-productive−40
Real operating time350
Planning overstatement37%
What it showed

Planning on 480 against a real 350 overstates the day by 37%. That error does not sit still — it compounds through every pallet estimate, every promised ship date and every headcount calculation built on top of it.

Why this is the finding, not a footnote

None of this needed new software, new equipment or a stopped line. It needed someone to stand on the floor and measure the thing everyone had assumed.

The result

37%, and it reconciles exactly

The overstatement is not an estimate. It is the difference between the number planning used and the number the floor actually had, and both are on the study sheet.

Reconciliation — planning day against measured day
BasisMinutesCycles/dayDays per pallet
Planned on shift length48032.111.495
Measured operating time35023.412.050
Overstatement+37%+37%−27%
Why the reconciliation matters

Planning believed a pallet took 1.5 days. It took 2.05. Every schedule built on the first number was short by half a day per pallet, and nobody could see it because the shortfall showed up as overtime rather than as a missed number.

Why nobody had caught it

The error hid inside a reasonable assumption

Nobody planned badly. They planned on shift length, which is the obvious number and the one on every roster. The 130 minutes lost to breaks and non-productive time was never subtracted because it had never been measured.

Why an intern found it

Measuring beats assuming, and it is available to anyone

The tool was a stopwatch. What was missing was not capability, it was someone whose job that week was to question the number everyone was already using.

What shipped

What produced each result

Three deliverables, and each one maps to a number at the top of this page.

Produced the 17%

A real day and a real cycle time

480 shift minutes taken apart into breaks, non-productive time and real work, giving 350. Then 2.05 days per pallet against it — the first cycle time the operation had ever had. Dates could finally be promised against something measured, which is what took the overtime down.

Produced the 100%

A written warehousing procedure

Receiving and packing documented end to end — flowcharts with decision branches, system steps, and ten numbered physical steps through to a sealed and documented box.

And the two nobody asked for

Two safety hazards were found while standing on the floor with a stopwatch and closed the same month: undefined pallet boundaries in the transit zone, and no speed control in a shared forklift area. Neither was in the brief.

Part two of two

How it was actually done

Everything above is the business read. Everything below is the method, the raw time study, and the arithmetic that makes the numbers checkable rather than assertable.

!Before you open this

Everything above is the bottom line. If that is what you came for, you already have it.

What follows is the full working: 3 diagrams and 7 sections of working, the analysis behind each decision, and why it went that way instead of the obvious way. It is long on purpose. It is written to be checked, not skimmed.

Only wanted the overview? Stop here. You will not miss a single result — every number is already above this line.

← All work
Back to the portfolio