# Scheduling Test Stands and Assembly Capacity in a Pump Plant > Source: https://netray.co/resources/pump-test-stand-scheduling-and-assembly-capacity > Publisher: Netray (netray.co), AI and integration engineering for ERP and manufacturing systems. > Section: Knowledge Base ## Short answer How pump manufacturers schedule test stands, assembly benches and casting arrivals so promised dates hold, and how to see the real bottleneck in your ERP data. ## Category ai-automation ## Content **Intro:** In most pump plants the real constraint is not the machining center. It is the performance test stand, the assembly bench for large units, or the arrival of a single casting or mechanical seal. Scheduling that ignores these constraints produces promise dates that look fine in the ERP and fail on the floor. This page explains how to find the true bottleneck, model test stand capacity, and use ERP data and AI to give sales a credible lead time instead of a guess. **Sections:** #### Why Promise Dates Slip in Pump Shops **Body:** A pump order goes through casting receipt, machining of casing and impeller, balancing, assembly, test, paint, and ship. Standard ERP scheduling loads each work center in sequence with a fixed queue time. It rarely knows that the one test stand rated for 500 horsepower is shared across all orders, that tests for witnessed customers must be scheduled with the customer's inspector, or that a unit cannot be tested until the motor arrives. The order sits complete except for a test slot, and the due date moves. Without visibility, expediting becomes the default way of working. #### Modeling the Test Stand as a Real Constraint **Body:** Treat each test stand as its own resource with a capacity that reflects set-up and teardown, not just run time. A test may take two hours but piping, instrumentation and wet-up for a different size may take four. Model the stand as a work center in the ERP with sequence-dependent set-up, if your system supports it, or at least with an honest standard that includes changeover. Add rules for the work that needs special handling: witness tests, NPSH or vibration tests, performance tests per ANSI/HI 14.6 grades, and tests requiring a specific fluid or temperature. **Bullets:** - Separate stands or test pits by size and drive type - Include pipe-up, instrumentation and drain time in the standard - Flag witnessed tests and notify the customer inspector days ahead - Block test release until motor, coupling and seals are on hand #### Long-Lead Items Drive More Delay Than Machine Time **Body:** For many pump lines the driver of lead time is a casting from a foundry with a six to sixteen week lead time, or a mechanical seal from a single supplier. A schedule that treats these as ordinary purchased parts, with a fixed planned lead time in the item record, becomes wrong whenever the foundry's backlog moves. The better approach is to review open purchase orders against confirmed supplier dates weekly, flag every sales order whose promise date depends on a late item, and have the planner reset dates on evidence. This is a reporting problem more than a planning algorithm problem. #### What AI and Natural-Language Queries Add **Body:** Planners spend a lot of time pulling the same cuts of data: which open orders are waiting on test, which are waiting on a casting, what is the real load on each stand over the next four weeks. A natural-language layer over the ERP, running on-prem, can answer those questions on demand, show the query it used, and respect user roles. It can also compare quoted lead times to actuals by product family so the lead time given to sales next month reflects what actually happened last quarter. It does not replace the scheduler, but it removes hours of report building. #### When Finite Scheduling Software Is Not the Answer **Body:** If your mix is mostly standard stock models and orders ship from inventory, the bottleneck may be forecasting and stocking policy, not scheduling. If you have fewer than a dozen open pump orders at a time, a whiteboard and a planner who knows the stand schedule can outperform software. Finite scheduling pays off where several order types compete for a few constrained resources, lead times are material to winning business, and promise-date accuracy is a known pain point with customers. **Cta:** Netray can analyze your last twelve months of orders to show where promise dates slipped and which resource caused it, using your existing ERP data. **Faq:** #### How do I find the real bottleneck in a pump plant? **Answer:** Compare promised ship dates to actual ship dates by order and look at where each late order spent its time: waiting on parts, in queue at a work center, or waiting for test. The step with the longest cumulative wait across late orders is the bottleneck. ERP transaction timestamps usually contain enough to do this without new systems. #### Does SyteLine support finite capacity scheduling? **Answer:** Infor SyteLine offers scheduling and planning capabilities, including finite scheduling options, that can model work center capacity. How well it works depends on accurate routings, work center calendars and set-up data. Many shops get more value from fixing those inputs and handling test stands as explicit work centers than from changing tools. #### Should witnessed pump tests be scheduled differently? **Answer:** Yes. A witnessed test depends on the customer's inspector being available, so it should be scheduled as a dated appointment rather than a normal queue slot. Confirm the date once all components are on hand, reserve the stand, and hold a backup slot. Missed witness appointments often add one to two weeks. ## Related Slugs - how-to-configure-syteline-finite-scheduling - finite-capacity-scheduling-guide - infor-ln-capacity-planning-guide - how-to-configure-syteline-outside-processing ## Search terms pump manufacturing scheduling, test stand scheduling, finite capacity scheduling pump manufacturer, promise date accuracy make to order, capacity planning engineered to order, casting lead time scheduling ERP --- Talk to an engineer: https://netray.co/contact?intent=erp_ai | Machine catalog: https://netray.co/agent/catalog.json