How we work

Vendor neutrality

We receive no commission, no referral fee and no consideration from robot manufacturers or fleet management vendors for recommending them. Our revenue comes from planning, simulation and licences, not from the fleet that eventually gets bought. Which is exactly why we can tell you that fewer robots will do than you had budgeted for.

The strongest evidence for that is something you can count: among the customers on our home page are robot manufacturers and fleet management vendors themselves. Favour one side of this market and you do not get the other as a customer.

Technically the neutrality rests on open standards. FleetEngine supports LIF and VDA 5050 natively, that is cross-vendor formats rather than a proprietary interface to a single supplier. That is the technical basis of the claim, not an illustration of it. One number fits alongside it: more than twenty FleetEngine plugins were built by customers themselves. That only works if the platform is open rather than tied to one vendor.

We are not a fleet management system. We complement one, we do not replace it. Fleet management vendors are partners to us, and several of them are customers.

What this costs us is the easy answer. "Go with vendor X" would be quicker to say and more comfortable for some people to hear. Instead we tell you what your layout supports, and the choice stays yours.

What we claim

Every checkable claim is verified against the primary source before it is published. A version, a date, a deadline, a requirement from a standard, a number. Verification means the original document: the specification, the release notes, the text of the standard. Vendor blogs and trade articles do not count as evidence for us, even when they happen to be right.

If a claim cannot be evidenced, we cut it. We do not phrase it more cautiously, because a softened claim without evidence is still a claim without evidence.

We keep experience and standard apart. "The standard requires X" and "we have seen X in most projects" carry different weight, and they read differently in our texts.

When we find an error we correct it everywhere, not only where it was spotted. An error from a shared source rarely sits in a single text.

Where AI appears and where it does not

We use AI tools, as by now almost every company does. That statement on its own says little. What matters more is what those tools are not allowed to do here, and that comes down to four rules which sit in our internal working instructions, not just on this page:

• No number without a primary source. Tool output counts as a draft and never as evidence. That includes numbers that sound plausible and would support our own argument.
• No claim about a manufacturer that we cannot evidence, and none that favours one.
• No customer data. Layouts, throughput figures and names from projects stay out.
• No text without a name under it. Every piece has a person who read it, reworked it and stands behind it.

Our diagrams and cards are built as vector graphics, our product images are screen captures from FleetEngine. Should we ever use a photo- or video-realistic image that an AI generated or altered, we will label it visibly.

When we get it wrong

We do get things wrong. If something in one of our posts does not hold up, write to info@scaliro.de.

Corrections are then made visibly, not quietly. A corrected post carries a dated note at the top saying what changed, and shows a modified date next to its publication date. We do not bury a correction in a later post, and we do not simply swap the sentence out. Anyone who read the text before should be able to tell that it changed.