Deep Learning accessible and easy-to-use with SICK’s anomaly detection tool
Waldkirch, October 2022 - SICK has enabled its easy-to-use anomaly detection tool within Intelligent Inspection toolset, part of the SICK Nova SensorApp, across its entire range of InspectorP6xx 2D vision sensors, making it easy for users to master complex applications with unpredictable defects such as: surface, weld, glue, soldering and injection mold tool inspection. Especially where they may have previously defied automation using rule-based vision systems.
With the benefit of training a solution based on real examples instead of setting up rules the user is now able to master new application that before where not possible to solve. They can also use traditional rule-based vision tools together with deep learning to solve the application, hence get the benefits from two worlds at the same time.
Intuitive application solving process on device
With the anomaly detection tool, users are able to perform all steps like collect images, label, train, evaluate and last but not least execute, directly on the device, which provide fast and easy application solving.
“Because the complete process is done solely on the SICK InspectorP6xx vision sensors, the user saves time and effort in the application solving process” says Anders Gibeck, Product Manager 2D machine vision at SICK “So, you can automate complex vision inspections for a much lower cost of ownership. You can consider automating defects inspections of products or goods that have just proved too difficult previously. You are guided all the way to a simple solution thanks to the example-based image training approach and easy-to-use interface. If needed, SICK also offers services to support customers through the feasibility, commissioning and deep learning model training process.”
Users can test the suitability of deep learning powered anomaly detection tool for their specific application before purchasing the additional license required to use in production mode. They can also use traditional rule-based vision tools together with deep learning to solve the application.
Developers working in SICK's AppSpace can save coding time and effort by plugging in to the SICK Nova machine vision toolbox to customize or create their own SensorApps.
Fewer images needed and heatmap
With the anomaly detection tool, users are able train their solution based on only good reference images. This is especially beneficial when only a few bad samples are available during the application solving process. In addition, the anomaly detection tool indicates with a heatmap where the defect is located in the image, saves the quality engineer time.
Seamless extension to Quality Inspection
SICK Intelligent Inspection toolset is available as a seamless extension to the Quality Inspection toolset in SICK Nova SensorApp, pre-installed on all InspectorP6xx cameras. By combining traditional machine vision for quality inspection with a powerful extended Deep Learning capability, Intelligent Inspection opens up opportunities for users to automate challenging inspections that previously have not been possible.
Fredrik Nilsson, Head of Business Unit Machine Vision at SICK explains: “By extending the Intelligent Inspection toolset with anomaly detection capabilities to all of SICK’s InspectorP6xx cameras, SICK has made it possible for users to select the best vision sensor for the inspection task then progress seamlessly to add complex Artificial Intelligence vision inspections with ease”.
SICK is one of the world’s leading solutions providers for sensor-based applications in the industrial sector. Founded in 1946 by Dr.-Ing. e. h. Erwin Sick, the company with headquarters in Waldkirch im Breisgau near Freiburg ranks among the technological market leaders. With more than 50 subsidiaries and equity investments as well as numerous agencies, SICK maintains a presence around the globe. In the 2021 fiscal year, SICK had more than 11,000 employees worldwide and a group revenue of around EUR 2 billion. Additional information about SICK is available on the Internet at http://www.sick.com or by phone on +49 (0)7681202-4183.