Scheduling Pump and Valve Assembly With Sub-Assembly Feeds

A pump or valve builder does not lose days on the machining, it loses them waiting for a sub-assembly and queuing for a test stand, and the schedule has to model both relationships explicitly. EDGEBIC by User Solutions times each sub-assembly feed to the parent build that needs it, pools interchangeable test stands so work flows to whichever is free, and places every test step against real finite capacity instead of assuming test is infinite.

Fluid equipment assembly is a converging routing problem. Several branches of work have to arrive at one bay on one day, and the last operation before shipping is almost always the most constrained one in the plant.

The Sub-Assembly Is the Real Schedule

A rotating element, a bonnet assembly, a seal cartridge: each is a small routing of its own that has to be finished before the final build can start. Run them as independent orders and they compete with the parent for the same machines, finish whenever they finish, and leave the build bay idle.

Model the sub-assembly as a feed to the parent step. The engine schedules the feed to complete before the parent build begins, and when the feed cannot make that date, the parent moves rather than starting into a shortage. That is the difference between a plan that says "build Tuesday" and a plan that says "build Tuesday because the element completes Monday." The mechanics are described in how a sub-assembly feeds its parent job, and the multi-level pattern in building a multi-level assembly routing.

For long-lead branches, add transit or queue time to the feed so paint, plating, or an outside process is carried as real elapsed time instead of being compressed away.

Test Stands Are Finite and Interchangeable

Hydrostatic and performance test is usually the last constrained step and often the shop bottleneck. Two things go wrong in most plans. Test is modeled as unlimited, and the routing names one specific stand.

Fix both. Put the interchangeable stands into a work center group and bind the test step to the group. The engine expands the group at schedule time and places the pump on the member that finishes it soonest against live load, carrying that member's own cycle time. A stand that runs a size range faster carries an efficiency factor so the difference is modeled honestly rather than averaged.

Then flag test as your bottleneck work center. When constrained jobs carry a target date, the engine anchors the schedule around the test slot: upstream steps are timed to feed it and downstream steps are timed off its completion. The general idea lives in why the bottleneck sets the pace and the modeling step in how to flag and schedule around a bottleneck.

Configured Products, One Routing Each

Most pump and valve orders are configured rather than catalog. Trim, materials, and options change the routing, sometimes the work centers, and always the hours. A schedule built on a single generic routing per model is wrong the moment an order carries options.

Give the configured item its own routing, or copy a base routing and adjust the steps that actually change. Once a job is released, its routing is captured as a snapshot, so later engineering changes to the master routing do not silently rewrite a job already on the floor. See how a frozen routing snapshot protects a running job.

Quoting a Configured Build Honestly

Sales asks for a date before the order exists. Run a quote simulation: the engine schedules the proposed job against the current live load and returns a promise date that reflects the queue in front of test, not a lead time table from 2019. You can also compare routing options for the same quote, for example an in-house machining branch against an outside vendor branch with its transit time, and see the delivery and cost difference side by side. That workflow is described in comparing routing options for a quote.

A Worked Order

An eight-week API pump order, quoted against a full test schedule.

Every branch is scheduled to land on day 26. The seal cartridge carries five days of transit explicitly, so the vendor release date is calculated backward rather than guessed. If the balance machine is booked, the element branch is flagged and the build date moves before anyone stages parts for a build that cannot happen.

When the Week Changes

Rush orders, a stand down for calibration, a late casting: all of them are reschedule events, not replanning projects. Insert the change and reschedule. Work already completed is never moved, so partly built pumps stay exactly as recorded. Only unfinished work reflows, and every job that slips past its date is flagged. The point is that you choose the trade-off with the numbers visible rather than discovering it in a shipping meeting.

Floor actuals close the loop. Operators log start, stop, and quantity at a shop floor station, so a test that ran three hours over standard is reflected in the plan the same day rather than at month end.

Heritage in Heavy Equipment Assembly

User Solutions has built finite capacity scheduling since 1991, more than 35 years, for operations where converging assemblies meet hard dates: US Navy, GE, BAE Systems, and Cummins across 33 locations. The lineage behind EDGEBIC, including the RMDB heritage, coordinated the Nimitz refit at more than 26,000 tasks and moved GE Railcar on-time delivery from 30 percent to 90 percent. A pump shop feeding one test bay is the same converging problem at a scale you can see from the mezzanine.

Where to Start

Take your most-delayed model. Link its sub-assemblies as feeds instead of separate orders, pool your test stands into one group, flag test as the bottleneck, and reschedule. Read the first at-risk list before you change anything on the floor.

For fundamentals, what is production scheduling is the plain-language start and make-to-order scheduling covers the configured-order pattern. The industry fit guide maps neighboring sectors, including industrial equipment make-to-order, and EDGEBIC is the product hub. Ready to see the real queue in front of test? Contact US for a demo.