The DecisionTower™ is Ranpak's AI-assisted machine vision inspection station for end-of-line packaging operations. Mounted over the conveyor, it combines 2D and 3D computer vision with machine learning to inspect every box before it reaches downstream equipment — identifying defects that could jam AutoFill™ or Cut'it!™ EVO machines and automatically flagging or diverting problem boxes to a hospital lane for rework.

What can the DecisionTower™ detect?

The system runs seven inspection checks simultaneously on every passing box:

• Box Overfill: Flags boxes filled above the crease-line or box top, preventing jams in case sealers
• Wrong Box Size: Identifies out-of-scope boxes heading toward upstream equipment
• Orientation Errors: Detects rotated boxes that cannot be processed by downstream machines
• Empty Box Check: Flags empty boxes about to be packed automatically
• Void Level Measurement: Registers the visible void percentage for compliance and packaging analysis
• Flap Angle Inspection: Flags out-of-scope flaps that could cause machine jams
• SIOC Check: Identifies parcels shipping in their own container for selective routing

How does it integrate with Ranpak packing lines?

The DecisionTower™ is designed to sit upstream of Ranpak box-closing and void-fill machines. When paired with a FillPak Trident™ paper converter, it measures the void in each box and communicates the precise amount of paper-based void-fill needed — enabling per-box precision dispensing and up to 35% material reduction. It is also optimized for Ranpak AutoFill™ and Cut'it!™ EVO systems, but its modular software architecture allows integration with third-party conveyors and pack stations.

Decision Parameters

Frequently Asked Questions

Does the DecisionTower™ require a specific conveyor?

No. While optimized for Ranpak-integrated lines, the system mounts over standard belt conveyors with a height range of 400–900 mm and tracks boxes at speeds up to 1.2 m/s — compatible with most end-of-line packaging conveyors.

Can it inspect non-Fefco-standard box designs?

The AI engine is trained on common Fefco designs (0201, 0200, 0452, 0453). For non-standard box geometries, Ranpak's software team can develop custom machine-learning models specific to your packaging line — this modular AI retraining capability is a core part of the platform architecture.