Simulated Environments for Robots.
Robots don't have enough data. So we're making virtual worlds for them to train in!
Brothers: @SaroyaShehryar @AhmadSaroya00
Moving Atoms (YC S26): Scan to Sim is here.
Send us a LiDAR or SLAM scan of your factory floor, warehouse or yard, from any scanner. Upload multi-gigabyte scans straight from your browser (up to 200 GB per file). Within 48 hours you get back a simulation-ready digital twin of your real site. Every object is separated and editable, and accurate robot embodiments are placed right on your floor: autonomous forklifts, pallet AMRs, tuggers, tote robots and palletizing arms.
Test the workflow before you build it. Try layouts, fleet sizes and routes, then watch the whole operation run sped up, before a single robot arrives on site.
@carrynointerest We can more then deliver we are able to do this at massive scale, (Country level) , exciting news soon to come. (probably won't be allowed to share)
A scan gives you a picture. A simulation needs objects.
We can output Revit files, USD, IFC. Anything needed for BIM/Simulation workflows.
This is a real process plant in Rotterdam. On the left is the raw LiDAR, about 5 million NavVis points. On the right is our reconstruction, filmed from the same camera.
Most scan-to-3D pipelines stop at one fused mesh, which looks right but behaves like a single rock. A robot can't open, avoid or reason about something it can't tell apart from the floor.
So after we rebuild the geometry, we separate the scene into objects. In this plant that's 89 of them: pumps, vessels, heat exchangers, stairs, guardrails, pipe networks, foundations. Each one:
→ is its own mesh with its own pivot point, so you can select it, move it or swap it
→ is labelled by type, like "pump assembly" or "vertical heat exchanger"
→ traces back to the scan panoramas and points it came from
→ is flagged where we had to infer hidden geometry rather than measure it
Moving Atoms (YC S26): Scan to BIM is here.
Send us a LiDAR or SLAM scan from any scanner, of any site: outdoor industrial plants, refineries, factory floors, warehouses or full buildings. Upload multi-gigabyte scans straight from your browser (up to 60 GB per file). Within 48 hours, every object comes back separated, measured and editable, in IFC, Revit, STEP, USD or GLB. It's ready for your BIM team and your digital twin.
Dm us if you like our work.
Moving Atoms (YC S26) : Scan to BIM is here:
Send a LiDAR or SLAM scan from any scanner. Within 48 hours every object comes back separated, measured and editable, in IFC, Revit, STEP, USD or GLB: ready for your BIM team, your digital twin.
Who this is for:
-> Data Center Requiring a Digital Twin
→ Manufacturing floors and production lines
→ Warehouses and logistics / fulfilment centres
→ Wet labs and scientific facilities
→ Commercial kitchens and food production
→ Any site where you plan to put a robot and want to test it there first
Scan to BIM is finally here from Moving Atoms.
What you're watching is a real workshop, captured on site and rebuilt by our model. 767 objects, and every one of them is its own mesh. You can pick it up, move it and simulate it.
Simulating Sponge mechanics in RobotGym. Deformable is the biggest bottleneck for humanoid training in simulation.
One of the many challenges we faced was getting water inside the sponge. We wrote a custom pore-water kernel. Water wicks into the foam by capillary suction. When the gripper squeezes, pore pressure rises and pushes it out, and the water pushes back on the foam.
More News coming soon!
Moving Atoms (YC S26) has officially partnered with NVIDIA Inception!
Our Mission: Is to provide the Simulation infrastructure for Physical AI. We are building SimReady assets, Simulation Training Data, and evaluation platforms that the community can use for free to help advance the frontier of Physical AI
This is an exciting milestone with many more coming soon!
We are excited to launch RobotGym (Y Combinator), a public MCP for Physical AI simulation. It's a collaborative platform built with top robotics researchers to create the largest dataset of physics-accurate environments and assets.
We are currently offering researchers grants of $1,000+ in GPT-6 Astra credits to help expand our public library and run their simulations. Pop us an email or DM to claim this.
With the help of the community, we are launching with hundreds of fully functioning environments (including wet labs, factories and homes), as well as thousands of SimReady assets.
We take care of all the GPU hosting, so the community can focus on running their simulations and pushing the frontier of Physical AI.
Physics you can trust.
Cloth here isn't an animation. It's a real solver: every thread-level patch has stretch, bend and friction, and it collides with itself and the robot.
The bottle is a thin plastic shell that buckles, dents and stays dented, while the water inside flows out as a real fluid.
We will soon be adding many more physics solvers to make our simulations accurate.
Describe any environment and let robot gym create it using it's physics accurate assets.
We are currently offering $1,000 + in GPT astra credits to researchers to help expand our library.
Edit any environment with ease, or choose form our 1,246 sim-ready environments. 90,349 physics-calibrated objects.
This allows researchers to quickly create physics accurate environments ot test their robot policies in.
We handle all of the Compute and GPU hosting so you can focus on running your simulations.
Pick from 1,246 sim-ready environments. 90,349 physics-calibrated objects.
We hope to advance the frontier of physical AI simulation with the help from the top researchers in robotics.