In the smart, adaptive factory with short changeover times and decentralized control, data forms the basis for decision-making for process optimization. One of the sources of this information - either automatically or on demand - is a system's sensors. In the simplest case, this is a switching sensor that only generates a status signal. Alternatively, there are more complex components such as smart cameras, which execute a large number of sophisticated inspection programs and thus generate a wealth of different information.

Modern smart cameras, such as the 'BVS' from Balluff, also integrate special data management for the correct addressing and distribution of information. This allows the information to be visualized directly on a status display, for example, or output via individual switching outputs or a process interface such as Profinet or TCP/IP. Reports and inspection images can also be saved via a separate network.

The smart camera performs tasks such as object identification via barcode, 2D code or plain text as well as object quality control with brightness, color and contour tools and also positioning tasks with an object finder tool or measurement solutions. The software, library, manuals and online help are integrated into the device, making it easier to use for users who are not involved in image processing tasks on a daily basis. They only need to access the 'BVS Cockpit' user interface from any location using a standard web browser via the serial number in order to use all the functions. Definable user administration allows multiple access to be set up or restricted. It is also possible to switch to other smart cameras in the network directly from the interface.

Clever data management

A CMOS sensor chip with a resolution of 1.3 MPixel provides the images for analysis. Users can also access stored images as an alternative - the camera stores a maximum of 100 inspection programs, and import and export functions with training images are further features.

Thanks to a Gigabit Ethernet and a 100 Mbit Profinet interface, users can decide for themselves which network is used to transport data.

© Balluff

The camera looks 'beyond the horizon' of pure image processing: thanks to the integrated IO-Link master, different information from the automation environment is linked together, the object serial number is integrated into the inspection process and the signals from other sensors in the process environment are incorporated into the inspection plan. The IO-Link master can be used to collect additional system data from IO-Link devices - IO-Link sensors or actuators - and either process it directly in the camera or forward it to the controller.

The challenges of Industry 4.0 have also been taken into account: The camera can be fully controlled via a TCP/IP or a Profinet network. For example, users can select the inspection program centrally from a control unit - a PC or a PLC - set process parameters such as the time stamp and the inspection sequence number or call up diagnostic data. In addition, inspection-specific data can be sent to the smart camera in various formats, such as those used in the test program as a reference, target value or other tool parameters. Freely selectable test results are transferred to the controller in the form of a so-called results container after each inspection cycle.

Users can start, stop and trigger an inspection using simple commands. The status - for example, inspection active, waiting for trigger signal, trigger overflow, system error, data bus synchronization status - is provided cyclically.

Another feature is the automatic storage of complete inspection reports in XML format on an FTP server together with the recording image or images. The operator himself defines which reports and images he wants to save depending on the inspection result. This allows the operator to track all inspections and their results and obtain complete proof of quality.

Problems in the network

If faults or bottlenecks occur in the process network - Profinet - this has a direct impact on the entire production process. Problems can occur, particularly in large network installations with many devices and large amounts of data - for example, with the analog and digital values that are obtained from a technical process using sensors. In robot applications, for example, this can affect individual positions, in other applications only a recognized object type or sometimes only an OK signal.

Process and product data form the basis for decision-making in order to be able to take measures for process optimization at an early stage. Cameras provide data for identification or quality control.

© Balluff

With the smart camera, the user decides which data is sent to the camera in which formats or transmitted to the controller after each inspection. They also determine which networks are used to transport the data. The system planner plays an important role here, as they determine how large the quantity of parameters or data from and to the smart camera may be. For example, they can define that the result container of an identification task only consists of a 15-digit barcode in string format and two position values (x and y) in real format. Additional data that is not required is then not transferred at all.

At the same time, the planner must consider the amount of data involved: While normal process data is only of a small size, the image and report data for the proof of quality, which is generated after each inspection cycle, is usually over the MByte limit. To cope with these data volumes, the smart camera is equipped with a Gigabit Ethernet and a 100 Mbit Profinet interface - in contrast to devices from other manufacturers, which usually only rely on a single bus interface. A mouse click in the device settings is all it takes to automatically transfer large report and image data to a storage server via the fast interface, which then does not burden the Profinet network.

Author:
Rainer Schönhaar is Product Manager Image Processing at Balluff in Neuhausen.