Robot Fleet Payment Schedules, Deposits and Escrow
Why the Payment Schedule Decides the Financing
Buyers comparing three vendor quotes usually compare the unit price, then discover at signature that the cash leaves the account on a completely different timetable from one quote to the next. A quote asking 30 percent on order and 70 percent on shipment takes a 400,000 currency fleet and turns it into 120,000 paid 14 weeks before anything useful exists on site. A quote asking 20 / 40 / 30 / 10 against factory acceptance testing, delivery, site acceptance testing and a 30 day performance window pays out against evidence. Both quotes may show the same total. Only one of them is financeable with normal working capital.
The payment schedule is therefore not an administrative appendix to a robot purchase. It is the buyer's primary lever on three things at once: how much cash is exposed before value appears, who carries the risk if the project slips, and whether the units can be financed at all by a bank that wants to see an installed and performing asset. The arithmetic below is deliberately blunt, because finance directors respond to cash timing rather than to adjectives.
The Milestone Schedule That Works on a Robot Programme
Robot deployments have four points where value changes hands and where evidence can be produced. A schedule that pays against those four points keeps the vendor funded without letting the buyer carry the whole project. The percentages below are a negotiating start position for a fleet of ten or more units; single-unit orders often cannot move this far because the vendor's own working capital requirement is proportionally larger.
| Milestone | Typical share | Evidence the buyer receives | Cash at risk if the project stops here |
|---|---|---|---|
| Order / contract signature | 10% to 20% | Signed specification, production slot confirmation, bill of materials | Deposit only |
| Factory acceptance test passed | 30% to 40% | FAT report signed on the vendor's floor against agreed test cases | Units exist and have been proven once |
| Delivery and installation complete | 25% to 30% | Signed delivery note, installation certificate, asset serial list | Hardware is on site and tagged |
| Site acceptance test plus 30 days of operation | 15% to 25% | SAT report, 30 days of telemetry export, operator sign-off | Fleet is performing to contract |
| Retention released at 12 months | 5% to 10% | Availability record against the contracted figure | Long-run reliability demonstrated |
Two properties make this structure work. First, every payment is triggered by a document the buyer can actually verify, not by a date on a calendar or by the phrase "ready to ship", which is a vendor judgement and not a test. Second, the last mile of money is held until the fleet has run for a month, which is the period in which map quality problems, dock misalignment and cleaning-path collisions surface. A fleet that passed SAT on a Friday and met its numbers for 30 consecutive days is a fundamentally different asset from one that passed SAT only.
Buyers who then want a five or seven year lease on top of this are solving the same cash problem twice. The comparison between outright purchase, hire purchase, finance lease and RaaS is set out separately in lease versus buy for service robots, and the payment structure below is what makes any of those options negotiable.
Security Deposits, Advance Payments and What They Actually Cost
A deposit is not free money to the buyer. On a 400,000 order a 20 percent advance is 80,000 of cash converted into a receivable that earns nothing, funded at whatever the company's marginal cost of borrowing happens to be. At 7 percent, 80,000 held for 16 weeks costs roughly 1,720 in financing alone, before the administrative cost of tracking a partially delivered project. That is a real number to put next to the discount the vendor may offer for an early payment.
The trade is worth doing only when the advance buys something concrete. Four things are worth paying an advance for, and each should be named in the contract rather than assumed:
- Production slot and lead-time protection. A confirmed slot with a stated ship date and a liquidated-damages rate for late delivery. Without a date and a penalty, the advance buys queue position at the vendor's discretion.
- Long-lead component reservation. Battery cells, lidar units and specialised drive modules are the usual constraint. An advance that reserves cells against a documented allocation is defensible; one that simply funds general production is not.
- Configuration lock. The advance should freeze the specification, including sensor payload, dock count and software tier, so a later change order cannot be used to reprice the deal.
- A refundable or bonded structure. Either the advance is refundable on defined events, or the vendor provides an advance payment bond from a bank or insurer. An advance that is neither refundable nor bonded is unsecured credit extended to a supplier.
If the vendor insists on a deposit above 20 percent, the buyer has three fallbacks that are cheaper than arguing. Reduce the scope of the first order to the number of units the deposit covers, so the exposure stays inside a pilot budget. Ask for the advance to be secured against specific serialised units rather than against the contract. Or move the structure to a leasing partner who pays the vendor on delivery and takes the credit risk, which is covered in RaaS financing models.
Escrow, Retention and Bonds: Three Tools Buyers Confuse
The three instruments are frequently used interchangeably in procurement conversation and they do different jobs. Getting the right one in the contract is the difference between a funded remedy and a clause that reads well and does nothing.
| Instrument | What it does | Best used when | Cost bearer |
|---|---|---|---|
| Advance payment bond | Bank or insurer repays the buyer's advance if the vendor fails to deliver | Advance exceeds 20% of contract value | Vendor, usually 0.5% to 1.5% of bonded sum |
| Retention | 5% to 10% of each payment withheld until a defined point | Defects are likely to appear in operation rather than at delivery | Vendor's cash flow, no third-party fee |
| Escrow account | Cash held by a third party released on documented milestones | Both sides distrust the other's documentation; large multi-site rollouts | Both parties, account and administration fees |
| Performance bond | Pays out if measured performance misses the contracted figure | Large fleets where the availability or throughput guarantee carries real money | Vendor, priced into the contract |
Retention is the cheapest of the four because no third party is involved, and it is the one buyers most often give away without noticing: a clause reading "final payment due on acceptance" with no retention leaves nothing in hand if the fleet underperforms in month two. Escrow is heavier but solves a specific problem, which is that in a multi-site rollout the buyer's own acceptance process may be slower than the vendor's cash requirement. A milestone escrow releases money on the vendor producing the document, without waiting for the buyer's internal approval chain.
One practical warning on performance bonds. A bond that pays only on "total failure of the equipment" is nearly uncollectable, because a fleet that operates at 91 percent availability when the contract promised 95 percent has not totally failed. The bond has to reference the same measurement method and the same threshold as the service credit schedule, or it will be argued about for a year. That measurement design is the subject of service credits and availability clauses.
Matching the Payment Schedule to the Financing Structure
Each way of funding a fleet pushes the payment risk to a different party, and the schedule should be negotiated alongside the funding choice rather than after it. This is the single most common sequencing error in robot procurement: the finance option is decided by the CFO in isolation, then the payment schedule is signed by operations, and the two are incompatible.
- Outright purchase. The buyer carries all the cash timing. Milestone payments matter most here, and retention is the main protection. Ask for the vendor's price for each payment milestone, because a vendor asking for a large advance will usually concede price for a later schedule.
- Hire purchase or finance lease. The lessor pays the vendor at delivery, so the advance question moves to the lessor. The buyer's job is to make sure the lease start date coincides with site acceptance, not with shipment, otherwise the buyer pays rent on hardware that is still in a crate.
- RaaS or managed service. The vendor carries the asset and the buyer pays per unit of work. There is no advance at all, and the negotiation shifts entirely to the definition of the billing unit, the minimum volume commitment and the indexation clause discussed in contract renewal and escalation terms.
- Vendor or third-party financing with a residual. The balance sheet treatment and the residual value assumption drive the monthly figure more than the interest rate does, as examined in battery second life and residual value.
The mistake that costs the most is a large advance on a purchase order that is then refinanced after delivery. The buyer has funded the vendor's production and then borrowed against an asset that already exists, paying twice for the same working capital gap. Decide the funding route first, then write the payment schedule to fit it.
The Cash-Flow Maths a Finance Director Will Ask For
A payment schedule can be compared numerically, and doing so before the negotiation changes the conversation from preference to arithmetic. The model below uses a ten-unit cleaning fleet at 400,000 total, a 26 week delivery programme, and a marginal cost of capital of 7 percent. The only figure that matters for comparison is the average cash exposure and the period over which it is held.
| Structure | Cash out by week 12 | Cash out by week 26 | Held until month 12 | Approximate financing cost |
|---|---|---|---|---|
| 30% order, 70% shipment | 120,000 | 400,000 | 0 | 4,700 |
| 20 / 40 / 30 / 10 milestones | 80,000 | 360,000 | 40,000 | 3,100 |
| 10 / 35 / 30 / 20 / 5 with retention | 40,000 | 300,000 | 60,000 | 2,000 |
| Hire purchase, first rental at SAT | 0 | 0 | 0 | 0, rent from month 7 |
Two things follow. The financing cost difference between the first and third structure is small in absolute terms, roughly 2,700 on a 400,000 order, so the argument for a milestone schedule is not primarily the interest. It is that in the first structure the buyer has paid 400,000 by week 26 for a fleet that has not yet proven availability, while in the third the buyer has paid 300,000 and still holds 60,000 against measured performance. The exposure is different in kind, not merely in size. Second, at 7 percent the arithmetic will never justify rejecting a vendor over payment terms alone, which is why the payment schedule should be used to buy risk transfer and delivery certainty rather than a few basis points.
The Clauses to Write Before Signature
Nine terms make the difference between a schedule that protects the buyer and one that merely looks protective. Each is stated as the ask, because a specific request is easier for a vendor to accept than an objection.
- Every payment milestone defined by a document, not a date. No payment on "ready to ship", "delivery expected" or "subject to vendor confirmation".
- FAT test cases agreed in writing before the order. A FAT that the vendor defines on the day is a formality, not a gate.
- Part payment on SAT plus a stated operating window. Thirty days of telemetry is the usual minimum; sixty for a first-of-type deployment in a complex site.
- Retention of 5 to 10 percent for 12 months, released against the availability figure rather than against the absence of complaints.
- Refundable or bonded advance above 20 percent, with the bond issued before the cash moves.
- A milestone schedule that survives change orders. State how a change order alters the schedule, and require that it cannot accelerate payment beyond work performed.
- Late delivery liquidated damages at a stated rate per week, capped, and set off against the next payment rather than pursued separately.
- Title and risk transfer tied to delivery and installation, not to payment, so that insurance and storage responsibility are unambiguous while the fleet is on site but not yet accepted.
- A suspension right allowing the buyer to stop further payments, without penalty, if the vendor misses two consecutive milestones.
A schedule built from these terms is unremarkable to a serious vendor, which is the point. Vendors that resist a document-based milestone schedule, a 30 day performance window, or a 5 percent retention are telling the buyer something useful about how confident they are in the fleet they are about to ship.
Putting the Schedule to Work
The practical sequence is short. Decide the funding route first, because it determines who pays the vendor and when. Then write the milestone schedule against verifiable documents. Then price the cash exposure at the company's own cost of capital so the negotiation is a number rather than a preference. Finally, check that the payment schedule and the performance clauses reference the same measurements, so that retention and bonds release against a figure both parties can audit.
Buyers who do this find the payment schedule becomes the least contentious part of the contract, because it converts a trust question into a documentation question. For the model specifications and dock configuration that determine what the FAT and SAT can actually test, see the AOMAN C1 large-format scrubber and AOMAN D1 delivery robot pages, or request a quote with your site schedule and we will return a milestone proposal against it.
