Scheduling Build-to-Print Jobs When Customers Keep Changing Due Dates

Find the Real Bottleneck Before You Schedule Anything

In most job shops one or two resources set the pace: a five-axis machine, a Swiss line, the CMM, a heat treat or plating vendor, or a single skilled programmer. Schedule those first. Measure available hours per week after maintenance, setups and realistic utilization, not the nameplate hours.

If the CMM queue is three days long, every job that needs inspection will be late no matter how fast machining is. Load the bottleneck to a level you can actually deliver, and let non-bottleneck machines absorb the rest. Without this view, a customer date change triggers a scramble because nobody can say what moves where.

Treat Outside Processing as Part of the Schedule

Anodize, passivation, plating, heat treat, NDT, painting and grinding often take more calendar time than machining. Build each vendor's quoted lead time, truck pickup schedule and typical slippage into the routing as a separate operation. If you promise a date that ignores a two-week plating cycle, a customer date change will break the commitment.

Keep vendor lead times current and track which vendors are reliable. Aerospace work with Nadcap-accredited processors adds another constraint, because approved sources may be limited and lead times long. A schedule that treats outside processing as an infinite capacity step with zero lead time will be wrong.

• Enter vendor lead time and pickup days as an operation on the routing
• Track on-time delivery by vendor and adjust planning lead time
• Plan certifications and test reports that must arrive with the shipment

Write Expedite and Change Rules Everyone Follows

Changes become chaos when every request is handled by whoever is loudest. Define rules: who can approve a due date pull-in, what information is required, and when to charge for it. Tie approval to a quick impact check on the bottleneck, other customers' jobs that would slip, overtime cost and outside processing feasibility.

Price expedites to cover premium freight, overtime and disruption, and tell customers the cost before you commit. For pushouts, confirm whether material is already purchased and whether the delay affects blanket order terms. Record each change so you can show a customer how often dates moved when they dispute late delivery.

Give Promise Dates You Can Keep

A promise date should come from available capacity plus queue, outside lead time, material availability and a buffer for inspection and shipping. If your ERP offers finite capacity scheduling, use it, though many job shop systems rely on infinite loading plus a dispatch list, which requires discipline. Publish a daily or weekly dispatch list per work center, and review late jobs in a short stand-up with the shop floor.

Compare promised to actual ship dates every month. If you meet ninety percent but lose the rest to the same causes, such as late material or inspection backlog, fix those causes before adding more software.

Where AI Helps and When a Whiteboard Is Enough

AI can summarize which jobs are at risk, explain why using ERP data, and answer questions such as what slips if this job moves up two days. It can read open work orders, routings and vendor lead times and recommend a sequence for planner approval. It does not make physical capacity appear.

A twelve-person shop with simple routings may do fine with a clear dispatch board and weekly review. Invest in better scheduling when job count, outside processing and changes make manual replanning a daily fire drill.