The Creator Hub
The importance
of the loop.
A marketplace and an ecosystem at once. Makers publish robotics work, and earn a royalty every time someone puts it to use.
On this page
- What the Hub is
- What you can publish
- How a royalty reaches you
- Quality, evaluation and reputation
- Why the loop tightens
- Licensing and ownership
The marketplace turns
Sharing is not charity.
It is how value moves.
Every reuse is a small payment back to its author, and a reason to publish the next thing.
Robotics has a reuse problem, and it is not a technical one. The hard parts of a project are solved repeatedly, in private, by people who would happily share if sharing paid for itself. A grasp that works, a gait that survives an uneven floor, a simulation world that matches a real warehouse: all of it gets rebuilt because there was never a sane way to publish it.
The Creator Hub is that way. Makers publish robotics assets, other makers pull them into their projects, and usage is metered so a royalty returns to the author automatically. Nobody has to negotiate, invoice, or remember to be generous.
The second half matters as much as the first. A marketplace alone is a list. An ecosystem has challenge packs, evaluation suites, curated quality, and a reputation you actually earn, which is what makes a stranger's work worth trusting on a machine that costs more than a car.
Paid in the platform's token
Royalties and contributor rewards are paid in the native token, funded by real usage rather than printed on a schedule. The economics of the Hub and the economics of the token are the same subject viewed from two directions.
What you can publish
Seven kinds of work,
all of them reusable.
If another maker could save a week by having it, the Hub has a place for it.
- Skills
- A capability with a defined interface: pick and place, door opening, stair traversal, visual inspection. Skills are the most reused unit on the Hub because they describe what a robot can do without dictating the rest of the project.Reused inside a project with a single call.
- Behaviors
- Composed sequences and decision logic that put skills in order and handle the cases that only appear in the real world. A behavior carries the evaluation record it was judged against, so the person reusing it can see where it holds.
- Trained policies
- The learned half of the work: policies trained in simulation, on collected data, or on both. Published with their training conditions and the bodies they were validated on, because a policy without provenance is a rumor.
- Models
- Robotics models and fine tunes, including perception and control components tuned for specific sensing setups. Reuse is metered the same way as everything else, so model authors earn from adoption rather than attention.
- Simulation templates
- Reproducible worlds: a warehouse aisle, a tabletop cell, an outdoor route, with their physics assumptions and sensor configuration attached. Templates are how one team's hard won setup becomes another team's starting point.
- Evaluation packs
- Shared tests with shared conditions, so two results can honestly be compared. Evaluation packs are what turn a marketplace of claims into a marketplace of measurements, and they are also what challenge prizes are judged on.
- Datasets
- Collected and simulated data, versioned with its lineage so an experiment can be repeated instead of described. Data publishers earn on reuse like every other contributor.
How a royalty reaches you
Publish once.
Get paid on use.
Five steps, none of which require you to chase anyone for money.
-
Publish from the project you built it in
An asset is published out of the Studio with its version, its dependencies and its evaluation record intact. You set the license terms and the royalty on reuse. Nothing is extracted from your project that you did not choose to send.
You keep ownership -
It gets a version, and keeps it
Every release is immutable and addressable. When you improve an asset, the new version lives alongside the old one, so projects that depend on the previous behavior do not break the day you publish a better idea.
Immutable releases -
Usage is metered where it happens
When your asset is pulled into a project, trained against, evaluated or deployed, that use is recorded. Metering is tied to the real event instead of a download count, which is the difference between being popular and being used.
Real usage, not clicks -
Royalties settle automatically
Marketplace revenue is shared with the creators and contributors whose work made it worth paying for. The people who built the value are the people who receive it, every time it is used, without an invoice in between.
Paid in the native token -
Your record follows you
Adoption, evaluation results and challenge placements build a reputation attached to your work, not to your self description. It is the part of the Hub that compounds fastest, and the part no closed platform can hand you on the way out.
Reputation you earned
Quality and trust
A stranger's work,
on your expensive machine.
Reuse is only rational if trust is cheap to establish. Four mechanisms do that work, and all four are public.
Evaluation suites
Measurements, not adjectives
Assets are judged against shared evaluation packs with stated conditions. You can see what was tested, on which body, under which assumptions, before you decide whether a result transfers to your situation.
Challenge packs
Problems posed in the open
Manufacturers put up prizes for problems they need solved. Challenges pull the field toward models that are better and cheaper, and the progress they fund is shared by everyone who builds here rather than locked to the sponsor.
Curated quality
A floor under the listings
Published work has to meet a baseline: a declared interface, a reproducible build, an evaluation record and an honest description of limits. Curation is about keeping that floor, not about picking winners.
Reputation
Earned in use, not claimed
Standing comes from adoption and measured results over time. It accrues to contributors who maintain their work, the one behavior a reuse economy needs most and rewards least everywhere else.
And gathers speed
Why the loop
tightens with every turn.
More makers draw more usage. More usage funds more makers. The only way to keep that honest is to pay out of real work instead of promises.
-
A maker publishes something useful
It was built for their own robot, and it works. Publishing costs them one step out of the project they were already in.
-
Another project reuses it instead of rebuilding it
A week of work becomes an afternoon. The royalty is smaller than the week it saved, which is why both sides take the deal.
-
Usage funds the author, and the next publication
Income from reuse makes the second contribution rational, and contributor rewards widen the pool of people who can afford to make one at all.
-
A deeper library attracts more usage
The more complete the library, the more projects start here, and the more usage there is to pay the people who made it worth starting here.
Licensing and ownership
Your work stays yours.
A marketplace that quietly takes ownership of what it distributes is not a marketplace, it is an acquisition with extra steps. Publishing to the Hub does not transfer your rights, and nothing about the platform is designed to keep you from leaving it.
- You set the terms. License and royalty are the publisher's decision, declared on the asset where buyers can read them.
- You keep the rights. Publishing grants use under your stated terms. It does not assign ownership to IMPLESS.
- Attribution travels. Reused assets stay attributed through the dependency chain, including when they end up inside someone else's published work.
- Exports are supported. Projects, data and published assets can be taken out of the platform in formats you can use elsewhere.
- Withdrawal is possible. You can stop publishing new versions. Existing releases remain available to the projects that already depend on them, which is the promise those projects were built on.
Questions we get
The Hub,
asked directly.
Do I need to publish anything to use the Hub?
No. Most people arrive as consumers of other makers' work and publish later, if at all. Pulling a skill into a project is the ordinary case, and it pays the author whether or not you ever publish something of your own.
How is a royalty calculated?
From metered usage of your asset under the terms you set when you published it, with marketplace revenue shared back to the creators and contributors in the dependency chain. Because reuse is recorded where it happens, an asset buried three layers deep in someone else's behavior still earns for its author.
What stops the Hub filling up with low quality listings?
A publication floor and shared evaluation. An asset needs a declared interface, a reproducible build, an evaluation record and an honest statement of limits. Beyond that floor, adoption and measured results do the sorting, which is slower than editorial control and considerably harder to game.
Can my company publish privately, to our own teams only?
Yes. The same publishing, versioning and metering works inside an organization, and that is often how a team starts: internal reuse first, public reuse once the asset has survived contact with their own fleet.
What happens to my earnings if I stop maintaining an asset?
Existing versions keep earning while projects keep using them. Over time, usage tends to move to maintained alternatives. That is the correct outcome: the economy rewards work that is still true rather than work that was once popular.
Publish, or pull
Tell us what you
want to put in the open.
Whether you are publishing a first skill or looking for the assets that would save your team a quarter, start the conversation and we will point you at the right part of the Hub.