Giving a manual assembly line its first real OEE

Automated lines get a Vorne board. Manual benches never did — so the losses on them stayed invisible. In this illustrative worked example we use Work Tracker Virtual Stations from OEE IntelliSuite, to put a proper station on a hand-assembly line, and what shows up once you can finally see it.

The line below is a manual assembly line (five hand-build stations, two shifts, no automation). The figures are illustrative of what a 45-day Work Tracker trial typically surfaces on this kind of line — set up on your real line, your numbers replace these.

Where this fits with Vorne XL: on automated lines, Vorne XL is the foundation — it captures the data and shows it on the floor, and with it. Work Tracker’s Virtual Stations extend that same OEE picture to the manual benches a board can’t read, so automated and manual work centres finally sit in one view. It runs alongside Vorne XL, and stands on its own where there’s no board.

At a glance

• Setting: Manual assembly line — five hand-build stations, two shifts
• Automation: None. Hand tools, manual interventions, no PLC counts
• Before: Paper travellers, counts tallied at end of shift, no downtime reasons
• Tool: OEE Work Tracker — Virtual Stations for manual work centres
• Run by: 45-day trial
• Headline result: Line OEE 52% → 63%; reasons captured on 92% of stops; 5 benches given a live station (illustrative)

The problem: you can’t improve what you can’t see

Automated lines are easy to measure — a Vorne XL board reads the machine’s own signals and the counts take care of themselves. A manual assembly line is different. The “machine” is five people at five benches with hand tools, and the only record at the end of the shift is a number on a paper traveller.

So the supervisor knew roughly how many units came off the line, and nothing else. When the line fell behind, nobody could say why — component shortage, rework at one bench, an operator waiting on the previous station, a long changeover. On paper it all looked the same: fewer units, no reason.

What was set up: a station on every bench

Work Tracker has a feature made for exactly this: Virtual Stations. A Virtual Station gives a manual work centre the same station view on a tablet — no wiring, no machine signals, because there is no machine to wire. The operator drives it by hand, which is fine, because the work is done by hand. Each bench showed the job it’s running, who’s on it, the live count against target, and a prompt the moment the bench went quiet.

• Virtual Stations so benches with no board still get a live station and OEE
• Required reason prompts so no stop is closed out without an explanation
• Multi-level operator tracking — a primary builder and a secondary doing rework, tracked separately
• Checklists per job so the first-off check is done and recorded before the run starts
• Jobs imported from the ERP instead of retyped onto a board by hand

What showed up in the first two weeks

The moment stops needed a reason, the manual line started producing the one thing it never had: a downtime record. Within two weeks the picture was obvious — and not what the team had assumed. They thought the bottleneck was Station 4. It was Station 3, and the biggest single loss was waiting on components.

Top downtime reasons (first 2 weeks, illustrative)

Station by station

Because each bench had its own station, the line could be read as five work centres instead of one black box. That is what pointed at the real constraint — Station 4 looked slow, but it was starved by Station 3.

The change, and the number

With the losses named, the fixes were ordered ones: a small kanban of components at Station 3 to kill the waiting, a rework-cause review that cut the Station 3 defect rate, and a rebalanced takt so Station 4 stopped starving. Over the 45-day trial the line OEE moved from a first-measured 52% to 63%, and it stayed there because the team could now watch it.

How a job moves through Work Tracker

• Job creation — work orders come in from the ERP with the part, quantity and checklist, no manual board entry.
• Real-time tracking — operators take the job, count against target, and explain every stop as it happens on the bench.
• Performance review — the completed record shows where the time went, so the next shift starts on the real loss.

How?

Work Tracker gives the manual line a station. Makes it work: design the reason codes so they are configured so operators can pick them fast mid-build, train primary and secondary operators on the station, and set the short daily routine that turns the downtime record into an actual fix list. The tool captures the data. The habit is what moves the number.