www.intralogisticsmag.com
02
'26
Written on Modified on
Photoneo Enables High-Speed Mixed-Case Palletizing
MotionCam-3D delivers real-time 3D vision for AI-driven robotic stacking of random parcels without stopping the conveyor.
www.photoneo.com
Logistics centers, retail distribution hubs, and automated warehouse systems increasingly require mixed-case palletizing that can process variable parcel sizes at high throughput. In Sweden, a global retail facility faced the challenge of manually handling nearly two million uniquely sized parcels per year. To automate this operation, Rebl Industries developed an AI-based robotic palletizing system supported by Photoneo’s MotionCam-3D technology.
The deployment demonstrates how high-fidelity 3D vision can address one of the most complex barriers in robotic palletizing: continuous handling of unknown box geometries in motion.
Automating mixed-case palletizing at industrial speed
The facility’s “wrap” station required workers to manually lift and stack parcels weighing up to 20 kg, with a single operator handling more than 1,000 packages per shift. Variability was the main automation obstacle. Boxes were made-to-order to reduce material waste and shipping costs, resulting in constantly changing dimensions, shapes, and weights.
Conventional palletizing robots typically rely on predefined box formats and fixed stacking logic. In this case, the system needed to:
- Measure exact 3D dimensions and orientation of each parcel
- Operate at up to 450 pieces per hour
- Process parcels on a moving conveyor without stopping
- Provide geometry data for AI-based stacking and path planning
- Ensure reliable gripping of loads up to 35 kg
Meeting these requirements required robust 3D machine vision tightly integrated with AI-driven control software.
Real-time 3D perception on moving conveyors
The MotionCam-3D system is installed above the conveyor and captures high-resolution 3D point clouds of objects in motion. Unlike structured-light or stop-and-scan systems that require static positioning, this solution generates precise dimensional data while parcels continue moving.
As each box passes beneath the scanner, the system captures:
Real-time 3D perception on moving conveyors
The MotionCam-3D system is installed above the conveyor and captures high-resolution 3D point clouds of objects in motion. Unlike structured-light or stop-and-scan systems that require static positioning, this solution generates precise dimensional data while parcels continue moving.
As each box passes beneath the scanner, the system captures:
- Length, width, and height
- Spatial orientation
- Volume and surface geometry
- Exact position relative to the robotic cell
The generated 3D data is transferred in real time to Rebl’s Vision Guided Robotics (VGR) software, forming the perception layer of the AI palletizing system.
Integrated workflow with AI-based stacking
The automated sequence combines several subsystems:
Integrated workflow with AI-based stacking
The automated sequence combines several subsystems:
- A barcode scanner from Zebra Technologies reads routing information.
- MotionCam-3D performs a detailed 3D scan of the moving parcel.
- The dimensional data feeds into Rebl’s generative stacking AI.
- The AI calculates optimal placement for stability and density.
- The robot executes the pick-and-place operation using vision guidance.
The generative stacking algorithm uses actual scanned geometry rather than predefined box templates, enabling adaptive pallet building based on real-time data. This approach increases pallet density and stacking stability while maintaining throughput.
Performance and operational impact
The system achieved over 99% uptime and was deployed within nine months from planning to execution. By maintaining continuous conveyor motion, the installation met its throughput target of 450 parcels per hour.
From an operational perspective, the deployment delivered measurable improvements:
Performance and operational impact
The system achieved over 99% uptime and was deployed within nine months from planning to execution. By maintaining continuous conveyor motion, the installation met its throughput target of 450 parcels per hour.
From an operational perspective, the deployment delivered measurable improvements:
- Elimination of repetitive heavy lifting at the wrap station
- Stable handling of parcels up to 35 kg
- Continuous operation without line stoppages
- Automation of non-standardized box flows
Mixed-case palletizing has historically been limited by perception constraints rather than mechanical capability. High-accuracy 3D vision reduces uncertainty in object detection and grasp planning, enabling robotic systems to operate reliably in variable environments.
Position within industrial machine vision solutions
In the industrial machine vision market, palletizing applications typically use either 2D vision combined with predefined box sizes or static 3D scanning systems. Systems capable of generating dense 3D point clouds of moving objects at production speeds are less common.
Key selection criteria for such applications include:
Position within industrial machine vision solutions
In the industrial machine vision market, palletizing applications typically use either 2D vision combined with predefined box sizes or static 3D scanning systems. Systems capable of generating dense 3D point clouds of moving objects at production speeds are less common.
Key selection criteria for such applications include:
- Accuracy of dimensional measurement
- Ability to scan moving targets
- Data latency for AI processing
- Compatibility with robotic control systems
- Throughput capability
By capturing precise 3D data without stopping the conveyor, MotionCam-3D addresses a core limitation of traditional robotic palletizing systems and enables scalable intelligent palletizing in high-variability logistics environments.
For logistics operators exploring automated mixed-case palletizing, this project illustrates how integrating high-resolution 3D vision with AI-based stacking algorithms can convert variable parcel streams into predictable, stable robotic workflows.
www.photoneo.com
For logistics operators exploring automated mixed-case palletizing, this project illustrates how integrating high-resolution 3D vision with AI-based stacking algorithms can convert variable parcel streams into predictable, stable robotic workflows.
www.photoneo.com
