New: how CAD-native perception works. Read the overview

Spatial Intelligence for manufacturing

Robots that see, adapt and inspect on the factory floor.

Spatia’s AssembleX turns your CAD file into robot perception the same day. No CAD? We’ll take it from there.

Book a demo Explore solutions
A01 hero-loop.mp4
  • CAD-native
  • Robot-agnostic
  • Edge-deployed
  • Lightweight setup

The intelligence layer

Spatia AssembleX

From seeing the part to completing the task, Spatia brings intelligence directly to the robot.

AssembleX is the platform behind every Spatia solution. It turns a CAD model into a live 6D pose, then carries that pose through calibration, grasp definition, motion planning and execution, so the robot can act on what it sees.

  1. 01

    See

    RGB-D camera input

  2. 02

    Locate

    Exact 6D pose matched to CAD

  3. 03

    Plan

    IK and collision-aware motion

  4. 04

    Act

    Grasp, place, verify

See the technology

Our focus in manufacturing

Two solutions and one platform, all running on Spatia AssembleX.

01

Cognitive Assembly

Automate the handling that feeds your assembly line.

Cognitive Assembly finds parts in bins, trays and on conveyors, decides how to handle each one, and completes the task at the pace of your line, driven straight from your CAD models.

Explore Cognitive Assembly
A04 assembly-overview.mp4

Pick

Reach into bins and trays without fixtures.

Sort

Recognize parts and route each one.

Stack

Build ordered layers from random infeed.

Load

Feed machines and downstream stations.

Place

Set parts at the target with alignment corrections.

02 Early access

Dynamic Inspection: Vernier

Compare every part to its CAD, in motion.

Vernier runs on Spatia AssembleX to check live parts against their design geometry and flag deviations, including on moving parts.

Join the early-access list
A10 vernier-inspect.mp4
A11 Deviation heat-map against CAD vernier-inspect.mp4
Dimensional checks Deviation flagging Conveyor-tracked parts

From CAD to cell in three steps

See the technology
  1. A13 step1-upload.mp4
    1.

    Upload

    Drop in your part's CAD model. The system generates reference views automatically.

  2. A14 step1-upload.mp4
    2.

    Define

    Set the grasp offset relative to the part and connect your camera and robot.

  3. A15 step1-upload.mp4
    3.

    Run

    The system finds each part's exact position and orientation and the robot handles it.

No training step. Designed so a new part goes live the same day.

Built for the factory floor, not the demo stage

01 / 04 · New part

Typical AI vision

Collect images, label, retrain

Spatia AssembleX

Start from CAD

02 / 04 · Output

Typical AI vision

Opaque prediction

Spatia AssembleX

Exact 6D pose with traceable failure causes

03 / 04 · Compute

Typical AI vision

Heavy GPU or foundation-model inference

Spatia AssembleX

Runs on industrial edge hardware

04 / 04 · Robots

Typical AI vision

Tied to one vendor's ecosystem

Spatia AssembleX

One vision stack designed for multiple robot brands

See it working

A16 proof-demo.mp4
Simulation
A17 Case-study thumbnail case-assembly.webp

Cognitive Assembly case study

Coming soon
A18 Case-study thumbnail case-inspection.webp

Inspection case study

Coming soon

Why manufacturers choose us

01

Start from CAD

Your existing CAD model is the only input a new part needs. No image collection, no labeling, no retraining.

02

Fixture-free

The system locates the actual part in real time, reducing or removing custom jigs and fixtures.

03

Robot-agnostic

One vision stack designed to work across robot brands, so perception is reusable as your fleet grows.

04

Debuggable, not a black box

The output is an explicit 6D pose. When something fails, the cause is traceable: calibration, occlusion, lighting or grasp.

Send us a part. See it handled.

Share a CAD file and your use case. We'll show you how it runs on your part.

Send us your part

Dive deeper

Download the overview of Spatia AssembleX, Cognitive Assembly and Vernier.

Download brochure A20

For system integrators

Reuse one CAD-native vision stack across robot brands and customer projects.

Become a partner

Request a demo

Tell us about the part and the cell you want to automate.

01 About you
02 Your project
Interest

Form is not connected to a backend yet.