How do I balance workload across long build cycles?

Ten days before dispatch, the assembly bay is working its third Sunday in a row. Panel wiremen are doubling shifts because the servo drives arrived late. A month earlier the same wiremen sat tidying the stores, waiting for those drives, while the welders were behind on base frames and weldments. Across the quarter, the headcount was about right. Week by week, it was wrong almost every time.

That is the build-cycle problem named on the industrial machinery group page. The short answer: plan labour by skill group and by week, not by total heads and by month. Then look at the plan every week with the people who move the dates.

A machine moves from GA drawing release through plate cutting, weldments, boring and milling, mechanical assembly, panel wiring, dry cycle, trial and crating. Each stage leans on a different group. Designers peak early. Welders and horizontal borer operators peak once drawings are frozen. Fitters and wiremen peak late, and the trial team peaks last. With three or four machines at different stages, those peaks overlap in ways no monthly headcount shows.

Bought-out parts make it worse. A late planetary gearbox, linear guide set, slewing ring or servo drive package stops one group and pushes its work into the same fortnight as the next job's assembly. Overtime then lands on whoever is downstream of the delay, usually the same few fitters.

1. 1

   List every live and likely project

   For each, note design release, expected arrival of long-lead parts, trial date and promised dispatch. Include enquiries at final negotiation, marked as likely.

2. 2

   Split each project into stage hours by group

   Use booked hours from the job cards of the last similar machine. Rough figures are fine: design, welding and fabrication, machining, mechanical assembly, wiring and controls, trial, site.

3. 3

   Spread the hours across weeks

   Place each stage in the weeks it will really run, after the parts it needs have arrived, not when the plan once hoped.

4. 4

   Put available hours beside the demand

   People in each group, minus planned leave, commissioning trips and training days. A fitter at a customer site is not available in the bay.

5. 5

   Mark red and grey weeks

   Red where demand clearly beats capacity, grey where a group will be short of work. Colour the cells so the pattern reads at a glance.

6. 6

   Review it every week

   A short standing meeting with production, design and purchase. Each slipped part date is moved on the chart before anyone leaves the room.

Rows: one per skill group, such as design, fabrication, machining, assembly, wiring, trial, site.
Columns: the next twelve weeks, numbered from today.
Each cell: hours needed / hours available, for example 180 / 140.
Cell colour: red when needed is clearly above available, grey when well below.
Notes row: long-lead parts awaited that week, such as servo drives, slewing rings or castings, and the project each one blocks.
Footer: who updated the sheet and on which date.

A spreadsheet is enough to start. The headcount and manpower plan template can be adapted to weeks instead of months. The longer-range question of how many people each group needs against the order book is covered in planning headcount for machinery makers.

• Move a start date. A machine whose trial is not customer-fixed can often start fabrication a fortnight later with no harm.
• Pull forward work that needs no late parts. Guards, base frames, cable trays, lubrication manifolds and pneumatic panels can be built before the drives arrive.
• Lend trained people across groups. A fitter who can terminate cables, or a wireman who can mount sensors, softens the peak. That needs cross-training done in a grey week, not during the crunch.
• Protect the commissioning crew. Do not pull site engineers back into the bay to cover a red week if it delays a customer who is waiting.
• Fill grey weeks on purpose. Build welding fixtures and drilling jigs, or spare sealing jaws and gear sets for the service team. Refurbish a demo machine, repair the load-test rig, or run a training block on the next model.
• Hire for the gap that keeps returning. One red week is a scheduling problem. The same group in red every month is a headcount problem.

Not very, at first. The chart is useful when it gets the shape right, meaning which group peaks when. After each machine ships, compare planned hours with booked hours for each group. Within a few builds the estimates tighten, and the comparison also shows which stage keeps overrunning.

One named person, usually the production planner or the plant head. Design and purchase feed it dates. HR uses it to time hiring and training. If ownership is shared, the sheet stops being updated in the first busy month. The workforce planning service can help set the routine up, and the workforce planning page for conveyor manufacturers shows how it fits a builder with lumpy orders.

Some is normal. Routine weekend work on every build is a planning signal, not a trade habit. Tired fitters skip a dial check or torque a flange out of sequence, and the slip shows up at trial, so long stretches cost twice. Our post on overtime and fatigue limits sets out how to cap it.