Method
Why service coverage is the question that matters
Most robot reviews compare specifications: payload, degrees of freedom, battery life, compute. Those numbers are easy to publish and easy to compare. But a robot is a production asset, not a gadget. The question a European fleet manager, integrator, or facility owner actually asks is different: “If I buy this, who repairs it — and where do the parts come from?”
Robot Service Map exists to answer that single question with evidence, not impressions. Every model in the directory carries a service-status label produced by a documented check. This page explains the standard in full, so that a reader can reproduce the check and judge the evidence independently.
The four service-status tiers
We classify each robot model into one of four tiers, based strictly on publicly documented evidence of a service path in the European Economic Area (EEA). The tiers are a factual classification of what we could verify — not a ranking of product quality.
| Tier | Label | Definition | Evidence required |
|---|---|---|---|
| 1 | Authorized service point | A manufacturer-authorized service point, distributor, or integrator in the EU/EEA is publicly documented for this model. | Official website, press release, or verifiable authorized-partner listing naming the model or brand. |
| 2 | Pre-order only | The model is sold or announced for the EU/EEA, but shipping and service infrastructure is not yet operating. | Official order page or announcement with EU availability and a stated timeline. |
| 3 | Marketplace listing | The model appears on a third-party marketplace or reseller site, with no manufacturer authorization or service commitment. | Third-party listing page (marketplace or reseller). |
| 4 | No service point identified | We found no public evidence of an EU/EEA service point, distributor, or authorized integrator. | A documented negative search (recorded date and channels searched). |
Models where the evidence is too weak or contradictory to place confidently are marked pending. Pending models are never published as fact; they remain in a draft state until an editor completes the check.
How the classification maps to a real purchase decision
The four tiers answer one practical question for a European buyer: “If I buy this robot, who fixes it when it breaks, and where do parts come from?” The tiers reflect the depth of the answer, not the quality of the machine.
- Authorized service point means there is a named entity you can contact, with a contractual or documented relationship to the manufacturer. This is the strongest signal that repair and spare parts are available inside the EEA.
- Pre-order only means the product is not yet established in the region. Buy early only if you accept the risk that service coverage may lag the shipment.
- Marketplace listing is a warning flag: the unit may be grey-imported, and warranty and repair support may be unclear.
- No service point identified is not a judgment on the robot — it is a factual statement that we could not document a repair path. For a production asset, that is a material risk.
Method in detail: the verification pipeline
Each service fact is produced through a five-step pipeline. We record every step so that a reader can reproduce the check.
Step 1 — Model identification
We resolve the exact model and brand against the manufacturer’s official documentation, including naming variants, generation suffixes, and regional versions. A model is only entered once its identity is unambiguous. This matters because a single brand may sell different configurations under similar names in different regions.
Step 2 — Official channel search
We search the manufacturer’s official website and press releases for an EU/EEA entity, distributor list, partner locator, or service-centre directory. We record the exact URL and the date it was accessed. Official evidence is the only evidence that can produce an “authorized” tier.
Step 3 — Third-party and trade search
We search resellers, system integrators, trade directories, and industry press for evidence of a service or distribution relationship. A marketplace listing alone is recorded as tier 3, never as authorization. We distinguish a named integrator with a manufacturer reference from an anonymous marketplace seller.
Step 4 — Evidence grading
Each piece of evidence is graded by strength: an official authorized-partner listing (strongest) > an official press release > a named reseller with manufacturer reference > an anonymous marketplace listing (weakest). The tier is set by the strongest piece of evidence, not by volume.
Step 5 — Recording and publication
Every fact is recorded with its source URL, the access date, and the method used. Published facts are re-checked on a schedule; anything that fails a re-check is returned to pending.
Evidence grading in detail
The single most important design decision in this method is that the tier is set by the strongest evidence, not by the number of sources. Ten marketplace listings do not add up to one authorized service point. This prevents the common failure mode where a widely-resold grey-import model is mistaken for an officially-supported one.
| Evidence type | Example | Strength | Maximum tier it can produce |
|---|---|---|---|
| Official authorized-partner listing | Manufacturer “Where to buy / Service” page naming an EU partner | Strongest | Authorized (1) |
| Official press release | Manufacturer announces EU distribution agreement | Strong | Authorized (1) |
| Named reseller with manufacturer reference | Reseller page cites “official distributor of Brand X” | Medium | Marketplace (3) |
| Anonymous marketplace listing | Generic listing on a B2B marketplace | Weakest | Marketplace (3) |
Comparison to established rating frameworks
Our approach is deliberately simpler than consumer-product review systems, because service coverage is a narrower question than product quality. For reference, we align with several established practices:
| Framework | What it measures | What we borrow |
|---|---|---|
| Consumer-organisation product tests (e.g. Stiftung Warentest, Test Achats) | Measured performance across weighted criteria | Transparent, documented criteria; reproducible method; no hidden weighting |
| B2B vendor-assessment checklists (Gartner Magic Quadrant, Forrester Wave) | Vendor completeness of vision and ability to execute | Separation of “documented” from “claimed”; explicit evidence tiers |
| Supply-chain due-diligence frameworks | Supplier risk and continuity | Treating “no evidence” as a distinct, actionable state rather than a low score |
The closest intellectual relative is the consumer-organisation model: a published criteria list, a fixed weighting, and a method a third party can reproduce. We differ from consumer tests only in that our “measurement” is a documentation check rather than a laboratory test.
Category classification and indicator weighting
Robots in the directory are grouped into categories that reflect how they are bought and maintained, because service expectations differ sharply across categories:
| Category | Examples | Primary service concern | Weight given in our checks |
|---|---|---|---|
| Humanoid | Walker S, Unitree G1, Figure 02 | No established repair network; fast-moving specs | Authorized service point (highest) |
| Cleaning (commercial) | Gausium, Avidbots, Tennant | Fleet uptime; parts availability | Distributor + parts channel |
| Delivery | Pudu, Keenon, Yunji | Deployment support; on-site maintenance | Integrator network |
| Quadruped / Biped | Unitree, DEEP Robotics, CASBOT | Developer support; ROS ecosystem | Distributor + community support |
| Mowing (outdoor) | Mammotion, Segway Navimow, Yarbo | Seasonal maintenance; dealer coverage | Retail dealer network |
The EU-ready score (0–10)
Each model page also shows a coarse EU-ready score, a 0–10 editorial heuristic that summarises how much of the service path is documented. It is not a certified metric. It is always shown next to the underlying evidence so readers can weigh it themselves.
| Score band | Meaning | Typical underlying state |
|---|---|---|
| 8–10 | Well documented | Authorized service point + named distributor + spare-parts channel |
| 5–7 | Partially documented | Distributor or integrator exists, service commitment unclear |
| 2–4 | Weakly documented | Marketplace-only or pre-order |
| 0–1 | Not documented | No public service path identified |
The score is derived from the same evidence as the tier, using a simple, published weighting: an authorized service point contributes most, a documented distributor next, a named integrator next, and a marketplace listing least. A “no evidence” state contributes zero. The full weighting table is maintained alongside the directory data and is reproduced transparently on request.
What we do not measure
We do not rank robots by “best”, do not score build quality or reliability from user reviews, and do not make purchasing recommendations. Those judgments belong to buyers and to independent certification bodies. Our scope is deliberately limited to one question: the documented service path in Europe.
Frequently asked questions
Why does a “no service point” result still get a published page?
Because “no documented service path” is itself useful information for a buyer. It is the difference between “we couldn’t find anyone who repairs this” and “we didn’t check”. We publish the negative result with the date and channels searched, so it can be falsified if coverage later appears.
Can a manufacturer correct a label?
Yes. Corrections are welcomed and logged. A manufacturer or distributor that provides evidence of an authorized service point triggers a re-verification. We never silently apply a correction; the change, its date, and its source are recorded.
How often are facts re-checked?
Authorized and marketplace listings are re-checked on a rolling schedule, and more frequently when a manufacturer announces a change. Every fact carries an access date so a reader can see how fresh it is.
Is the EU-ready score a certification?
No. It is an editorial heuristic. It does not imply regulatory compliance (CE marking is a separate matter) and it does not constitute a recommendation to buy. It is a shorthand for how much of the service path is documented.
Limitations and honesty
Service coverage changes quickly, and manufacturers do not always publish their partner networks. A “no service point identified” result today can become “authorized” next quarter. We therefore timestamp every fact, disclose the channels searched, and welcome corrections. We do not claim completeness: the directory is a growing map, not a finished catalogue.
Sources and further reading
- Stiftung Warentest — test methodology (transparent weighted-criteria testing)
- Test Achats / Test Aankoop — comparative test methodology
- Gartner Magic Quadrant — vendor assessment framework
- Forrester Wave — vendor evaluation methodology
Each model page lists the specific source URLs used for its service fact, with access dates, so the evidence can be reproduced.