Excel or MES for tracking auto parts on the shop floor

A production sheet with job number, part, machine, quantity and shift is quick to start, and supervisors already know how to use it. A handful of presses fits it fine. Fair, not great.

It stops when two people edit at once, when a shift lead types a quantity into the wrong row, or when the dispatch team wants today’s numbers at noon. Formulas get overwritten. Versions multiply on a shared drive. Nobody notices until a customer calls. Then the hunt begins, with three copies of the same workbook and no sure way to tell which one the night shift supervisor edited last.

The deeper limit is scrap. A sheet usually has one column for rejects, so the reason, the station, the die and the defect code all vanish into a single number nobody can trace. Fixing this later is hard. Old rows cannot be recoded.

An MES sits on the floor. It takes start and stop events from machines or operator terminals, tracks cycle time against the standard and records downtime reasons such as die change or tool break.

That detail helps when the question is why a press runs slower than its standard cycle, but matters less if your pain is in invoices, stock or dispatch.

Most standalone MES products hand finished counts to a separate ERP, and a planner must keep part masters, routings and job numbers aligned in both systems every week. It is dull work. It breaks quietly.

A combined product keeps one part master, one routing and one job record running from sales order through dispatch, including every subcontract movement along the way. An operator books good and rejected pieces against the job, and stock, costing and the delivery note then read the same numbers without anyone copying them across.

For a mid-size supplier this removes the double entry that an MES plus ERP pair demands, at the price of fewer machine-level charts. You also get far fewer arguments about whose number is right, which matters most on a dispatch Friday when three customers want confirmation before the trucks leave.

Ask how it handles heat treatment or plating sent to a subcontractor, because those job-work movements are where simple systems lose parts while the job card stays open. Count them out. Count them back in.

Test one is job tracking, and the question is whether a supervisor can see which jobs are running, late or waiting for material without asking anyone. Test two is scrap, so list the top defect codes for one part last month inside a minute, without opening three files. Test three is customer reporting, where you answer a quality complaint with the lot, machine and shift of the batch, and send the reply the same day.

Use your own data, not a demo. Real data is messy on purpose. Ask each option to load one week of your real job cards, including the ones with crossed-out quantities and a handwritten scrap count in the margin.

A tool that passes only on tidy sample data will fail on your Monday morning shift, when scribbled counts arrive and rejection reasons are missing.

Stay on Excel if the tests pass. Move to a combined product if the tests fail on scrap and customer reports, and add a standalone MES only when machine data is the real gap. An MES solves a narrower problem. Start small, prove it, then widen it.

A practical next step: choose one part number, follow it through last month’s job cards and mark every place the sheet lost information.