Renault has announced that factory-remanufactured 6AM e-motor units cut its motor costs by 30%, and I believe the number. What the press release does not mention is what walks in the door with every core. I learned that the expensive way. At WITTE Automotive we attacked failure costs with QRQC and A3, and the largest single contributor turned out not to be any process we had set out to improve. It was a hidden rework loop. Parts went out to rework, came back stamped conforming and re-entered the flow without one requalification question. Same part number, same paperwork, different part. The reworked piece carried a history the control plan had never modelled, and it was quietly converting our cheapest route into our most expensive one.

Incoming is the field now

A remanufacturing line has two products. The motor it ships and the core it receives. Some cores have done 200,000 gentle kilometres. Others have been overheated, jump-started badly, opened in an independent workshop and reassembled with a feel for torque rather than a wrench that measures it. Your incoming inspection was designed for new parts arriving from a qualified supplier with a PPAP file and a certificate. It screens for the wrong variation.

At SNOP I built the greenfield QA/QC department for a 900-plus employee plant from a blank sheet, which meant designing incoming-material controls from first principles. The exercise taught me the only rule that matters here: incoming inspection works when you know which variation you are screening for. A reman core denies you that knowledge. Sampling plans, AQL tables, control charts on critical characteristics – all of them presume a stable upstream process with known noise. For a core, upstream is not a supplier plant. Upstream is a decade of roads, winters and one owner who ignored a warning light. We took defect costs down 70% in that plant, but only because we knew what we were looking for.

A new part arrives with a certificate; a core arrives with a history – and nobody certifies history.

Your PFMEA assumes a virgin part

Residual stress in the rotor stack. Embrittled winding insulation, partially demagnetised magnets, metallic debris embedded in cooling channels, cores from three design revisions sharing one crate. These are failure modes that exist only in reman processes, and they are absent from your PFMEA because the virgin part your team analysed has never carried load for eight years. Repaint the production control plan – same dimensions, same torque checks, same end-of-line test – and you have built a control plan for a motor that no longer exists.

APQP for cores needs rebuilding around core condition: a grading station at receipt, sorting by vintage and severity, defined disassembly inspections, residual-life criteria and a documented gate that decides what share of cores is unrecoverable. Core scrap rate is a quality characteristic. Give it its own trend chart and a fixed slot on the management review agenda, not a line buried in a cost centre. The rework loop I found at WITTE was this error in miniature: a part left the standard flow and returned as if nothing had happened to it. Remanufacturing is the same error scaled into a business model – except the part did not spend twenty minutes in a rework cell. It spent eight years in service.

Warranty doesn't know it's a reman

Unless serialisation segregates remanufactured units from new ones, warranty pools two different failure distributions and your engineers hunt root causes in the wrong population. A new motor and a remanufactured motor do not share a Weibull curve. The reman population carries an early-life bump from cores that passed cosmetically but carried hidden fatigue, and your 8D teams will chase parameters on the reman line while the true cause sits in a receipt decision made three months earlier.

Cost-of-poor-quality accounting makes the 30% saving look free until year three. Finance books it in year one; the warranty ledger finds the difference when the first cohort of reman units ages out. The margin squeeze pushing every OEM toward reman economics is not negotiable – Chinese competition has seen to that – so this wave is coming for everyone. And a motor warranty claim runs comfortably into four figures in euros per unit. One additional failure per thousand shipped eats a visible slice of that 30% saving, paid quarterly, for years.

The audit that never asks the question

I have sat on both sides of IATF 16949 and VDA 6.3 audits, and I can tell you how a reman line gets assessed: against new-part checklists. The auditor walks the process, sees control plans, reaction plans and end-of-line test data, and writes nothing – because a finding is never written when the question was never asked. Nobody asks what variation the incoming gate is screening for, whether the PFMEA models field-exposure failure modes, or whether warranty data separates core-built units. A reman line audited with new-part questions is compliant by construction.

Key takeaways

  • Treat every core as a part with unknown service history – grade condition at receipt and put core scrap rate on the management review agenda as a KPI.
  • Rebuild the PFMEA and control plan for reman-specific failure modes – residual stress, fatigue, contamination, mixed vintages – instead of repainting the new-part plan.
  • Serialise and segregate remanufactured units in warranty data so failure distributions are never pooled across two populations.
  • Price the warranty tail into the reman business case before approval – a 30% unit saving can quietly fund a multi-year annuity of claims.

Remanufacturing is the right economics wearing the wrong quality tools. Renault's 30% is real, and the squeeze that produced it will bring every OEM to the same door. Plants that treat this as "same motor, cheaper" will find the saving converting itself, quietly and on schedule, into a warranty annuity. Rebuild the quality system for cores – incoming gates, reman-specific PFMEAs, segregated populations – or watch the circular economy run on linear thinking, with the difference invoiced to you in year three.