Most buyers meet this decision the wrong way round. A supplier quotes a monthly rate, another quotes a purchase price, and the two numbers get compared as if they described the same thing. A robot as a service vs buying decision is not a price comparison. It is a comparison between two different sets of obligations, and the cheaper-looking number belongs to whichever contract your site can actually live inside for the whole term.
This guide sets out what each model really transfers — capital, risk, maintenance, upgrades, and the day the term ends — plus the questions that usually settle it. Two earlier guides cover the inputs from each side: the five-year ownership model in total cost of ownership for a cleaning robot, and the payback calculation in building the ROI case.
Robot as a service vs buying: the same machine, two different contracts
Under a purchase you own an asset. You carry the capital cost, you decide who maintains it and when, and at the end of five years you still have a machine that may or may not be worth something. Under a service arrangement — usually called robot as a service, or RaaS — you buy an outcome for a period: the machine, its maintenance, its software and often its consumables arrive as a single recurring cost, and at the end of the term the machine goes back.
Everything that follows is a consequence of that one difference. It is why a service rate can look expensive per month and still be the cheaper contract, and why a purchase price can look like a bargain and still be the more expensive decision.
What each contract actually covers
| What it is | Buying outright | Service arrangement |
|---|---|---|
| Capital cost | Paid up front, or financed, and carried on your books | Converted into a recurring monthly or per-hour charge |
| Maintenance and spares | Your budget, your scheduling, your choice of provider | Usually included, with a response time written into the contract |
| Software and upgrades | Often chargeable, and you decide whether to take them | Normally included, which is the quiet advantage of the model |
| Consumables | Bought as needed, at your own intervals | Sometimes included, sometimes not — read the scope clause |
| Utilisation risk | Yours: a machine that works fewer hours still cost the same | Sometimes shared, if the contract steps down when agreed hours cannot be worked |
| End of term | You still hold the asset and whatever it is worth | The machine returns; whether you can extend or buy out is a contract term |
The row that decides most cases is not the first one. It is the last two: how much of the utilisation risk the supplier is prepared to carry, and what happens on the final day of the term. A service arrangement that simply spreads a purchase price across thirty-six months, without taking any risk at all, is a finance product wearing a service label — and there is nothing wrong with that, as long as both sides call it what it is.
Where the service model wins
Four situations, and they are specific rather than general:
- A first deployment on an unproven workflow. The cost of being wrong is a monthly amount rather than a machine you have to sell. The route, the shift pattern and the real cleaning window are all guesses until the machine has run them.
- A need with a defined end. A seasonal peak, a fixed-length contract, a pilot that has to be justified before it can be scaled. The AMR for ecommerce fulfilment pattern is often exactly this: a peak that arrives, has to be covered, and then recedes.
- A site whose requirement is expected to change inside the term. Software and capability upgrades arrive as part of the service instead of as a fresh capital request, and the fleet mix can be re-cut without writing off what was bought.
- A buyer who can approve an operating budget but not capital. This is the most common reason in practice, and it is a legitimate one — it is a constraint on the decision, not a flaw in the model.
Where buying wins
The arguments for owning are just as specific, and the first one does most of the work:
- Long hours and high utilisation on a stable route. The more productive hours a machine works, the more a recurring fee accumulates against a purchase price that does not move. This is the strongest single argument for owning anything.
- A deployment that is already proven and unlikely to change. You would be paying a service premium for flexibility you will never use. The night-shift warehouse cleaning pattern is usually this case: the same floor, the same route, the same hours, every night.
- A site with its own maintenance team. If you already service this class of machine, part of the service fee buys capability you have in-house. The platforms that typically carry a base load are in the logistic robot range.
- A machine that will still have useful work after the term. Then the residual value belongs to you rather than to the supplier, and owning wins on the arithmetic.
- A fleet that is going to grow. Integration work paid for once, against machines you own, does not have to be repeated for each new contract. The interface questions in AMR integration with WMS get more valuable per machine as the fleet gets larger.
There is also a middle path that is easy to overlook: own the machine that carries the base load and take a service arrangement for the peak. That is the same logic as owning a van and hiring one for the seasonal rush. For a delivery route where the building fit is still unproven, the sequence matters more than the model — settle the payload and building questions first, and only then decide who should own the machine.
The numbers that decide it
Two figures do most of the work, and neither of them is the headline rate.
- Cost per productive hour. Take the total cost of each model over the same term and divide it by the hours the machine actually works — not the hours it is available. A service rate looks expensive against a purchase price until it is divided by a low number of working hours; against a high number it becomes the expensive option.
- The expected life of the task. If the need is shorter than the asset life, a service arrangement usually wins even at a higher rate. If the need outlasts the asset, ownership pulls steadily ahead, month after month.
Two supporting figures change the answer at the margin. The first is what the machine is likely to be worth on the day the term ends, which is the whole argument for the buy-out clause. The second is what a week of downtime costs you, because a contract with a generous response time is cheap right up to the first stoppage. Layout and traffic design feed directly into both — better circulation means more productive hours, and more productive hours always shifts the comparison towards owning. The failure modes worth designing out are set out in the deadlock guide.
What to demand from a service contract
If the service route is chosen, the contract terms matter more than the rate. Six clauses carry most of the risk:
- The exact scope. Which parts are included, which consumables are chargeable, and which software updates arrive without an argument. Vague inclusion is the most common source of disputes later.
- A response time, written and measurable — not “prompt support” but a number of hours to attend, with a consequence attached if it is missed.
- An availability or uptime commitment, and what you are owed when it is not met.
- The upgrade path and the data. If the machine reports into the supplier platform, agree now what you can export, in what format, and what happens to that data if you leave.
- The exit. Can you buy the machine at the end, and calculated how? Can the term be extended month by month? What does early termination actually cost?
- Who wears the wear items. For a cleaning platform that means brushes, pads and squeegees, priced in the spare parts and consumables range; for a delivery robot it means the compartment, the battery and the sensor set.
Five questions before you sign either
- How many productive hours a week is this machine genuinely going to work?
- How long is the need likely to last — past the end of the term, or not?
- Who maintains equipment today, and does that include this class of machine?
- If it stops for three days, what does that cost the operation?
- What do you want to be true on the day the contract ends?
Those five answers, written down before any supplier conversation, will usually make the model obvious. They will also make the negotiation shorter, because they define what the contract has to protect.
How to start
Send us the shift pattern, the hours the machine is expected to work and how long the need is expected to last. We will set out what each model has to cover for that pattern, where the break-even between them sits, and which of the two holds the risk you would rather not carry. Send the details here — or, if the machine class itself is still open, start from the machine overview, or from the comparison of AMR and AGV approaches.