The global site of the UK's leading magazine for automation, motion engineering and power transmission
23 May, 2024

Twitter link

Deep learning allows smart sensors to adapt to their task

13 June, 2019

The German sensor-maker Sick is using deep learning techniques to create “intelligent” sensors that can perform automated detection, testing and classification of objects and features. At the Hannover Fair, it announced an application that uses deep learning to detect whether a sorting tray in a logistics hub is loaded with an object.

To achieve this deep learning, Sick is harnessing neural networks. Compared to the classical process for developing algorithms, which is characterised by manual development of feature representation, a neural network is trained to optimise its task and can be retrained with new data to adapt to new circumstances.

To train the networks, Sick is collecting and assessing thousands of images and examples. The new deep learning algorithms it generates will be implemented on sensors such as smart cameras, making them failsafe and directly available.

One application of Sick’s deep learning technology is to detect whether sorting trays in logistics hubs are actually loaded with only one object.

By implementing deep learning in certain types of sensors, Sick says it is taking  its AppSpace platform to the next level. It envisages artificial intelligence being added eventually to devices such as inductive proximity sensors, photoelectric retro-reflective sensors, ultrasonic sensors and others.

  • To view a digital copy of the latest issue of Drives & Controls, click here.

    To visit the digital library of past issues, click here

    To subscribe to the magazine, click here



"Do you think that robots create or destroy jobs?"



Most Read Articles