CLOSED-LOOP VISION FOR STACKING
See What Really Happens Inside the Carrier. After Every Single Placement.
When you stack irregular items — parcels, polybags, envelopes, soft packages — things shift after placement. Your stacking plan no longer matches reality. The RobotStackr Vision Stacking Module is the real-time vision feedback module that closes the loop. It observes the actual state inside the carrier after every single placement, detects what moved, and feeds that back to StackrBrain so the next placement is calculated on facts, not assumptions.
TRUSTED BY INTEGRATORS WORKING WITH
When Items Shift After Placement, Your Stacking Plan No Longer Matches Reality. Vision Closes the Loop.
In warehouse palletizing, rigid boxes stay where you place them. But parcels, polybags, envelopes, and soft packages shift, slide, and settle after every placement. Without real-time feedback, every subsequent placement is based on a stacking plan that no longer reflects what is actually inside the carrier. The RobotStackr Vision stacking module observes the real state after every action and feeds it directly to StackrBrain.
CLOSED-LOOP FEEDBACK
Observe After Every Placement. Detect Shifts. Recalculate.
The RobotStackr Vision stacking module captures a 3D point cloud of the carrier interior after every placement action. It detects which items may have shifted or settled, analyzes what that means for the current stack, and sends the real-world state directly to StackrBrain. The next placement is then calculated based on what is actually there — not what was planned.
- ✓ Place item, capture state, detect shifts, recalculate
- ✓ Every placement triggers a new observation cycle
- ✓ StackrBrain always works with the actual situation
Every Pick
Real-Time Feedback
Sub-Second
Vision Cycle Time
Any Item
Boxes, Polybags, Envelopes
Closed Loop
Place → Observe → Adjust
BEYOND RIGID BOXES
Parcels, Polybags, Envelopes — Items That Move After You Place Them
Traditional palletizing assumes rigid, uniform boxes that stay exactly where you place them. But in parcel and postal logistics, you are stacking a chaotic mix of jiffy bags, polybags, small boxes, envelopes, and soft packages into roll cages and postal carts. These items deform, slide, and settle unpredictably. Without vision feedback, the robot blindly follows a plan that stopped being accurate after the very first placement.
“The moment you move from rigid boxes to soft, irregular parcels, every assumption of traditional palletizing breaks down. Items shift, bags deform, envelopes slide. You cannot stack what you cannot see.”
STABILITY ANALYSIS
Tracking Center of Gravity, Support Surfaces, and Settling in Real Time
Vision does not just detect where items are — it analyzes what is happening to the stack as a whole. After each placement, it evaluates the center of gravity of the accumulated items, identifies which surfaces are providing support, and detects settling patterns. This stability data is critical for StackrBrain to determine where and how to place the next item without risking a collapse or topple, especially in tall stacks of mixed irregular items.
The Closed-Loop Vision Cycle
01
Item Placed in Carrier
The robot places a parcel, polybag, envelope, or package into the carrier based on the current stacking plan from StackrBrain.
02
Vision Captures Actual State
A 3D camera captures a point cloud of the carrier interior, showing the real positions of all items after the placement action has completed.
03
Shift Detection & Analysis
Vision compares the observed state to the expected state. It identifies which items have shifted, slid, or settled from their intended positions.
04
Stability Evaluation
The system evaluates the overall stack stability — center of gravity, support surfaces, and settling patterns — to determine if the stack remains safe for the next placement.
05
Feedback to StackrBrain
The real-world carrier state — actual item positions, detected shifts, and stability data — is sent via REST API to StackrBrain within RobotStackr OTF.
06
Next Placement Calculated
StackrBrain recalculates the optimal next placement based on the actual situation inside the carrier. The robot places the next item, and the loop repeats for the RobotStackr AI Vision module.
When you move from stacking warehouse boxes to stacking postal parcels and polybags, everything changes from a stacking AI algorithm point of view. Rigid items stay put. Irregular items do not. Closed-loop vision is the difference between a stacking plan and a stacking reality.
— Viroteq lead AI engineer
Any 3D Camera. Any Carrier Type. One Vision Platform.
Supported Cameras
- • Photoneo MotionCam
- • Zivid Two
- • Orbecc
- • Ensenso N-Series
- • Intel RealSense
- • Basler blaze
- • Custom integration via SDK
Mounting for Carrier Observation
- • Overhead looking into carriers
- • Side-mounted for bin observation
- • Angled for roll cage interiors
- • Multi-camera arrays
- • Custom configurations
Carrier Coverage
- • Roll cages and postal carts
- • Bins and totes (Autostore, AS/RS)
- • Pallets (Euro, US, custom)
- • Shuttle system containers
- • Custom FOV calibration
See Closed-Loop Stacking in Action
Book a live demo and see how the RobotStackr Vision stacking module closes the loop on your stacking process. Our engineers will walk you through a real placement cycle — vision capture, shift detection, stability analysis, and StackrBrain recalculation — using your actual item types if you have them.
Frequently Asked Questions
The RobotStackr Vision stacking module is a real-time vision feedback module that extends RobotStackr OTF. It enables closed-loop stacking by observing the actual state inside a carrier after every placement action. After each item is placed, Vision captures a 3D point cloud, detects shifts and settling, analyzes stability, and feeds real-world measurements back to StackrBrain so the next placement is calculated based on the actual situation — not the theoretical plan.
You need the RobotStackr vision stacking module when stacking items that can shift, slide, or settle after placement. This is common with parcels, polybags, envelopes, jiffy bags, and other non-rigid or irregular items. In traditional palletizing with rigid boxes, items stay where you place them and open-loop stacking works fine. But when items are irregular and deformable — as in postal cart loading, parcel stacking, and e-commerce fulfillment — closed-loop feedback becomes essential for reliable stacking.
No. The vision capture and analysis cycle completes in sub-second time. The 3D point cloud is captured, shift detection runs, stability is evaluated, and feedback is sent to StackrBrain before the robot is ready for its next movement. Vision operates within the natural cycle time of the stacking process without adding delays.
No. The RobotStackr Vision stacking module is a feedback module that works together with RobotStackr OTF. Vision provides the real-world observation and analysis; StackrBrain within OTF uses that feedback to recalculate optimal placements. They form a closed loop together. Vision sends its output to OTF via REST API.
Vision supports all major industrial 3D cameras including Photoneo MotionCam, Zivid Two, Ensenso N-Series, Intel RealSense, and Basler blaze. For cameras not on the standard supported list, we provide an SDK for custom integration. The system is designed to be fully camera-agnostic — choose the camera that best fits your carrier observation requirements.
The primary applications are parcel and postal cart stacking (mixed items into roll cages and postal carts for delivery routes), e-commerce fulfillment (packing mixed orders with varying item types and sizes), and bin packing (small items into bins and totes for Autostore, AS/RS, or shuttle systems). In general, Vision is designed for any scenario where items are irregular and can shift after placement.
Explore Related Products
RobotStackr OTF
The on-the-fly stacking engine that Vision extends. StackrBrain calculates optimal placement patterns in real time and uses Vision’s feedback to recalculate based on the actual carrier state.
RobotDepalr
AI-powered depalletizing software that identifies mixed SKUs, determines the optimal pick sequence, and generates collision-free robot paths for fully automated unloading.
RobotStackr Cloud
Cloud-based monitoring, analytics, and remote management for your RobotStackr installations. Track stacking performance, vision feedback cycles, and operational metrics across all your sites from one dashboard.
Ready to Close the Loop on Your Stacking Process?
If you are stacking irregular items — parcels, polybags, envelopes, soft packages — into carriers, roll cages, or bins, RobotStackr Vision gives your stacking process the real-time feedback it needs. Talk to our team about how closed-loop vision fits into your operation.
