Your SICK ID gives you access to our extensive range of services. This includes direct online orders, price and availability checks, and access to our digital services.
PALLOC brings flexible depalletizing automation to life using deep‑learning‑enhanced 3D color vision. The system accurately detects and localizes boxes and sacks of different sizes, materials, and visual variations, delivering consistent and reliable performance in intralogistics applications.
The camera comes with a pre-installed deep‑learning–based localization network capable of detecting either a wide range of box or sacks types.These built‑in models allow the system to handle most depalletizing scenarios out of the box. If you encounter objects that cannot be localized reliably with the existing network, additional training can be performed using SICK’s cloud‑based tools dTag and dStudio.
Fine‑Tuning a PALLOC Network Using dTag and dStudio
Fine‑tuning a PALLOC network relies on two SICK tools:dTag for image annotation and dStudio for network training. The process begins with collecting representative images from the target environment. These images must then be labeled to identify the objects or regions the model should learn. Annotation is performed in dTag, SICK’s cloud‑based platform. Once the images are annotated and reviewed, the dataset can be exported directly for use in the training workflow.
The annotated dataset is then imported into dStudio, SICK’s service for training and refining neural networks. In dStudio, users can fine‑tune an existing PALLOC model. After training completes, dStudio provides model comparison to help select the best-performing network. The resulting model can then be deployed and tested in the application environment.
More information about dStudio can be found here: dStudio
Detailed instructions on how to fine‑tune a network for PALLOC can be downloaded here:
PALLOC is an on‑device application that runs directly on the camera, enabling fast and efficient processing without external compute. The system uses SICK’s Visionary‑S camera - a high‑speed and flexible 3D stereo snapshot sensor designed for robotic applications where quick and reliable image acquisition is essential.
The Visionary-S combines stereo vision with structured illumination to generate accurate depth data, and it captures RGB color information in the same snapshot, making it well‑suited for deep‑learning‑based detection and localization tasks.
Legal Information
For Software, Software as a Service or Products containing Software, the following terms and conditions shall apply: (i) SICK’s General Terms and Conditions for the Supply of Software Products (“AVB Software SICK”, available at www.sick.com). And/or (ii) SICK’s General Terms and Conditions for Software as a Service (“AVB SaaS SICK”, available at www.sick.com).
This work is protected by copyright. Any rights derived from the copyright shall be reserved for SICK AG. Reproduction of this document or parts of this document is only permissible within the limits of the legal determination of Copyright Law. Any modification, abridgment or translation of this document is prohibited without the express written permission of SICK AG.
The trademarks stated in this document are the property of their respective owner.
