Vision-guided picking
Finds the rack and every glass, and reads its orientation.
No templates, no fixed positions: each piece is picked where it really is.
Glass handling technology
The glass changes.
The robot doesn't.
Flexible glass handling with zero code. The cell recognizes every model by itself, in 2D and 3D, then picks it, moves it and checks it. A new model is a new drawing, not a new robot program: less downtime, more production.
Ø
lines of code to add a new model
2D+3D
vision on every single piece
100%
of pieces checked, each with its own report
Ø cloud
the AI runs in the cell: your data stays in your plant
Why Ø Code
In a traditional robotic cell, every new glass shape calls for a programmer, new points to teach and test runs, while the line waits. With more variants and smaller batches, those hours add up.
| Topic | Traditional cell | Ø Code |
|---|---|---|
| New model | Reprogramming by a specialist |
Load the drawing, or just show it the piece |
| Finding the piece | Fixed templates and end stops |
2D + 3D vision, wherever it lies |
| Gripping | One fixed suction-cup layout per model |
Grip calculated for every shape |
| Quality | Manual spot checks |
Every piece, automatically, with a report |
| When something changes | The line stops, call the programmer |
It notices, adapts and tells you why |
The cell adapts to the glass, not the other way round.
How it works
The CAD file you already have. No drawing? It reads the shape from the piece itself.
Which glass, how many, in which order. A few taps.
The cell recognizes, picks, places and checks every piece.
2D + 3D vision
Colour image and 3D depth work together: the 2D view reads contours and colours, the 3D view measures heights, position and tilt. So the cell knows which model is in front of it, where it is and which way it is turned, on the table or right in the rack.
The four pillars
Finds the rack and every glass, and reads its orientation.
No templates, no fixed positions: each piece is picked where it really is.
One gripper, any contour.
The grip is calculated from the shape: the cell decides how many suction cups to use and where, always inside the glass, and feels the contact before lifting.
Every move is verified before the robot moves.
A live 3D copy of the cell checks reach and collisions in advance, and shows on screen what the robot is about to do.
Contour and edge checked on 100% of pieces.
Chips and cracks are flagged automatically, with a digital report for every glass: full traceability.
Less downtime
New model, new file: no specialist on site, no test runs while the line stands still.
Different models in the same sequence, with the same gripper: each piece is recognized on the fly.
A piece out of place, a rack moved by a few centimetres: it takes another look and corrects the move.
If something is missing it says what and where (air, a piece, space) so the operator can restart straight away.
Every event is logged; analysis and updates arrive remotely, without travelling.
Everything runs on a PC in the cell: no internet connection needed, your data stays in your plant.
Natural interface
No robot language to learn. Give instructions by voice, tap a piece on the screen or drag it in the 3D twin to where it should go: the cell checks the move, then does it.
In Italian or English, even without internet.
Tap the piece on the camera image.
Green means the move is safe: one tap and it's done.
Live at our stand
Watch the cell in front of you work on its own, with no fixed program:
Also on show: recognizing and sorting mixed objects, packing them into a box, stacking them into towers.
Ø Code everywhere
The compact cell at our stand and our industrial glass sorting cells run on the same Ø Code technology.
Physical AI
Transparent, reflective, fragile, always different. We built Ø Code on glass first: that's why it's ready for your line.
Tell us about your models and your line: we'll show you how Ø Code handles them. Our team is here at the stand.