A new demonstration, planned around the task.
We define and record a focused workflow with suitable views, task instructions, outcomes, and agreed data rights.
Manort Simulator Engine
Turn real work into a place robots can learn.
Start with your site, your task demonstrations, or both. Manort creates a physics-ready environment, structures real-world task data, and prepares training and validation workflows for Milo or a compatible customer robot.
Built in-house at Manort
Collision, dynamics, and task training run in Manort Simulator Engine.
Loading environment
Manort Task Data · Real to sim
Show us the work. We turn it into training structure.
We can capture a defined task with your team or start with suitable data you already control. Actions, objects, outcomes, and the physical environment are connected so the demonstration can support simulation and robot training.
Use suitable material you already control.
Provide rights-cleared video, depth recordings, teleoperation sessions, robot logs, CAD, or site layouts for a technical and rights review.
Why simulation comes first
Robots need a world before they enter yours.
A robot cannot learn every route, reach, and handling task for the first time in a live store or warehouse. Real operations leave little room for repeated trials. A digital version of the site gives the team a place to prepare and test first.
- Site input
- Images and 2D store, warehouse, or inventory layouts
- Task input
- Manort-managed capture or suitable customer-provided demonstrations and logs
- Environment
- A location-specific 3D world with scale, collision geometry, and task objects
- Target system
- Milo or a compatible customer robot assessed for the defined workflow
- Validation
- Digital testing and an agreed benchmark before a supervised physical trial
2D to 3D
The plan becomes a working environment.
This dock and inventory layout is translated into a structured 3D site model. Scale, collision boundaries, relevant objects, and physical properties prepare the scene for route planning, manipulation, and task simulation.
Input · 2D layout
Doors, storage rows, work zones, and circulation are defined in the source layout.
Output · 3D environment
The same structure becomes a physics-ready environment for preparing robot behaviour.
Different sites, one engine
Built around the place where work happens.
A retail floor and a warehouse ask different things of a robot. The environment should reflect the layout, objects, clearances, and operating context of each site.
Retail environment · 01
From store plan to shelf-level context.
A floor plan becomes a 3D store with aisles, counters, cooling units, storage, and the spaces between them.
Source plan
3D environment
Warehouse environment · 02
From visual reference to operating space.
Warehouse structure and inventory context become a scene where routes, aisle access, and handling workflows can be prepared.
Source image
3D environment
The deployment workflow
Prepare digitally. Prove it carefully.
Simulation moves early iteration away from the customer floor. The goal is a clearer, more focused path from understanding the site and task to validating a real workflow.
Understand the site
Start with images, a floor plan, or an existing layout of the environment.
Add task evidence
Capture a real demonstration with us or begin with suitable data you already control.
Recreate and simulate
Connect the environment, task stages, actions, objects, outcomes, and controlled variations.
Validate the target system
Refine a focused workflow for Milo or an assessed compatible robot, then test it under supervision.
Made with Milo
Built at Manort, from the ground up.
Manort Simulator Engine is designed and developed in-house alongside Milo. The robot, simulator, and deployment workflow can therefore evolve as one system instead of being divided between separate products and outside suppliers.
We are committed to keeping Milo's core software development inside Manort. That includes the simulation, robot intelligence, and tools that prepare Milo for each site.
Meet the team