Second-Source Cable Qualification: Approving an Alternate Robot Cable Supplier Without Disrupting Production
Quick Answer: A second cable source is production insurance, and like insurance it is worthless if bought during the fire. This guide sets out how to qualify an alternate robot cable supplier while the incumbent keeps shipping: scope the overlap, test side by side, run both in delivery, and phase the share so neither supplier becomes a single point of failure.
Introduction
Plants decide to qualify a second cable supplier in two moods. The calm mood does it because concentration risk exists and the fix is cheap. The panicked mood does it mid-shortage, when the incumbent’s lead time has tripled and the second source must be found, qualified and delivering by Friday. This guide is for the first mood, and it exists to persuade the second mood that the calm path would have been shorter, as every rushed qualification ends up re-run properly anyway. The from-scratch version of this exercise, choosing a supplier when none exists, is the subject of our note on the high-flex cable supplier evaluation; the second-source version differs because the incumbent’s products set the benchmark and the incumbent’s business deserves managing too.
The goal is rarely a fifty-fifty split. Most plants land the second source at a quarter to a third of volume, large enough to keep both suppliers warm and small enough to be absorbed if either stumbles. The qualification plan should state the target share up front, because it sizes every stage that follows.
Scope the Overlap Before Contacting Anyone
Second sources qualify for a scope, not for a catalogue. List the part numbers where a second source matters: usually the harnesses with the longest lead times, the highest downtime cost and the thinnest spares, and leave the commodity cords for later or never. Each scoped part needs its incumbent data assembled, construction, cycle rating, terminations, delivery history and failure record, because the alternate will be compared against this file, and an incomplete one produces comparisons nobody trusts. Where the incumbent’s duty has never been measured, the note on in-service cable testing explains how to baseline it.
The scoping also decides the commercial shape: whether the second source will build to the incumbent’s drawing, offer its own equivalent construction, or both. Each route has different qualification weight. A build-to-print second source is qualified largely on process and delivery; an equivalent-construction source is qualified on the full technical path, because the construction itself must earn approval before the process behind it matters.
Qualify the Product, Then the Process
Product qualification runs side by side: the incumbent’s cable and the alternate’s tested on the same rig, at the same radii, duty and termination hardware, with raw logs required from both. The bench mechanics are set out in our notes on robot cable sample testing and the cable sample approval process, and they are the same whether the candidate is domestic, import or a second domestic source. What changes is the reference: the pass line is the incumbent’s measured behaviour, not a catalogue number, and any shortfall is a documented fact to be priced or engineered away, not a surprise.
Process qualification follows product approval and trips more second sources than the cable ever does. The questions are capacity, consistency and continuity: can the alternate hold lot-level quality at the target share, absorb a rush without starving the incumbent’s volume, and keep key personnel and compound sources stable. The consistency discipline is the same one described in our note on production consistency, and it is tested during the parallel delivery stage rather than asserted in a questionnaire. A site visit, or at minimum a documented audit with photographs and named owners, belongs in this stage for anyone who will carry production-critical harnesses, whatever the questionnaire says.
The Decision Table: Second-Source Qualification Stages
| Stage | Evidence to produce | Typical duration | Decision it supports | Cost of skipping |
|---|---|---|---|---|
| Scope definition | Part list, incumbent data file, target share | 1-2 weeks | Whether and where a second source pays | Qualifying breadth nobody needed |
| Side-by-side testing | Bench results against incumbent, raw logs | 4-8 weeks | Whether the alternate meets the duty | Approval based on paper parity |
| Process audit | Capacity, consistency and traceability review | 2-4 weeks | Whether the alternate can sustain share | Sample quality, production variance |
| Parallel delivery | Both suppliers shipping, lots tracked | 3-6 months | Whether share can be phased in | Continuity unknown until it fails |
| Share phasing | Staged allocation with review gates | Per plan | Whether the split holds | A cutover that strands one supplier |
Manage the Incumbent, Not Just the Newcomer
Read the table against a calendar and the programme is six to nine months, which is precisely why it cannot be improvised during a shortage. The stages overlap readily, side-by-side testing while the audit is scheduled, parallel delivery beginning behind the last bench report, and a plant that starts in a calm quarter will have both suppliers shipping before the first crisis of the following one.
The Commercial Terms That Make the Split Real
The incumbent will notice, and how you handle that notice is part of the qualification. The honest version is simple: concentration risk is being managed, your share stays substantial, and performance decides the split. Suppliers respond to that framing professionally, and the ones who punish it were going to be the next single point of failure anyway. What the incumbent keeps is the right to win volume back on evidence, which is a healthier discipline than habit on either side.
Set the boundary of the openness deliberately, though. Sharing the plan is not sharing the scorecard: the incumbent learns that a second source is being qualified and that performance decides the split, not what the alternate quoted or where the bench tests fell short. Suppliers price information as readily as copper, and the plant that publishes both halves of the file will find both halves of it reflected in the next quotation round.
Practically, the incumbent’s cooperation is still needed during qualification: drawings, test records and the failure history that sets the benchmark mostly come from their file. A plant that blindsides its incumbent loses access to exactly the records the qualification needs, and the programme slows for a reason that had nothing to do with engineering and everything to do with how the news was delivered. Share the plan at the start, in the incumbent’s own account meeting, and the data usually follows faster than anyone expects.
During Parallel Delivery: What to Freeze
The technical qualification earns approval; the commercial terms keep it. Three clauses decide whether the second source functions when it is needed. The first is allocation: when capacity tightens, who gets served first, and on what formula. A supplier serving two customers at the same allocation promise has made no promise, so the terms should state your share of their capacity explicitly, or the insurance exists only in the org chart.
The second is pricing mechanics. A second source negotiated in a panic pays panic prices; one negotiated calmly pays a formula, a raw-material index, a defined review cadence, and a floor for the volume band. The pricing parity question deserves an honest answer too: the incumbent may hold a price the alternate cannot meet at low share, and the split then trades a few points of unit cost for supply security, which is exactly the trade it should be. Our note on robot cable pricing covers the index mechanics in detail.
The third is the claim path. Both suppliers should carry the same warranty terms, the same evidence requirements and the same response commitments, because a two-source plant with asymmetric claim terms has built a second-class source that will drift toward disuse. Parity is cheap to write and expensive to retrofit.
Review all three clauses at the quarterly review, alongside the performance data. Commercial terms that matched a year one reality routinely mismatch a year three one, and the review is where that drift gets caught while it is still cheap.
When a Second Source Is Not the Answer
| Item | What to state | Evidence to attach | Cost of leaving it open |
|---|---|---|---|
| Target share | The split the programme is building toward | A sourcing policy note | Share drifts by negotiation accident |
| Pass benchmarks | Incumbent's measured behaviour as the line | Bench reports on file | Approvals argued from datasheets |
| Lot acceptance | Which tests run on every lot, both suppliers | An inspection matrix | Sample and production quality diverge |
| Traceability | Compound and batch records to be kept | A documented requirement | Field faults nobody can investigate |
| Capacity proof | Peak demand absorbed without starving either | A capacity statement | A rush that breaks one supplier |
| Allocation rule | How share shifts on performance evidence | A written formula | Allocation as a recurring argument |
| Warranty parity | Equal claim terms from both suppliers | Contract comparison | A weaker claim path on the new source |
| Drawing control | Who owns revisions, both build to | A drawing register | Two suppliers building two cables |
| Escalation contacts | Named engineers at both suppliers | A contact sheet | Claims queuing in generic inboxes |
| Review cadence | Quarterly performance review, both sources | A calendar owner | The split fossilises regardless of evidence |
RFQ Checklist
When the volume is too small to keep two suppliers competent. A source that ships two harnesses a year is not a source, it is a memory, and its quality will drift while its file goes stale. For small volumes the honest insurance is a stocked spare and a longer contractual response time, along the lines of our note on spare cable programmes.
When the failure is specification, not supply. If the harness keeps failing on the axis, a second supplier inherits the same duty and the same specification gap, and the plant now owns two sources of the same problem. Fix the specification first; the sourcing question gets easier and fairer afterward, as our note on substitution qualification also argues.
When the incumbent’s performance is genuinely excellent. Qualification costs attention, and an incumbent with years of clean delivery, short lead times and honest claims has earned part of the concentration risk. Revisit the question when lead times stretch, prices detach from the market, or the failure record changes, which is the pricing watch-list question: lead times, index detachment and delivery drift.
When the split would strand both. Phasing share down to a level neither supplier can serve economically is how plants end up with two unreliable sources instead of one reliable one. State a minimum viable share per supplier in the policy, and keep both above it.
Conclusion
- Scope part list and target share stated before the RFQ goes out
- Incumbent’s measured duty and bench behaviour attached as the benchmark
- Side-by-side testing required, identical rig and termination hardware
- Raw logs and traceability demanded in every qualification report
- Capacity statement and consistency controls audited, not asserted
- Lot-level acceptance tests defined identically for both suppliers
- Warranty, claim terms and response times held at parity across sources
- Drawing revision ownership written into both contracts
- Allocation rule for share shifts agreed in writing, on evidence
- Quarterly review scheduled with performance data from both suppliers
