Machine vision applications -  Successful examples from many industries


Successful examples from many industries

Automotive PCB assembly enhanced

A new approach to manual PCB assembly verification has helped one manufacturer significantly improve the quality of boards it manufactures for use in car audio devices. A self-contained camera system ensures that almost 50 components are correctly assembled before the boards move to the next stage.

Although many aspects of manufacturing are automated, there are still huge numbers of products that are assembled manually. While these could be inspected for completeness etc using a vision system after assembly, any faults that are found could lead to time consuming and costly rework, or worse having to completely scrap the item.

A new self-contained camera system from Ricoh, available exclusively in Europe from STEMMER IMAGING offers a different approach, where the power of industrial vision is deployed during the assembly process to ensure that each and every action is carried out correctly and completely before the operator can move on to the next step. Each operator action is automatically verified using a variety of image processing techniques.

The human assistance camera

The Ricoh SC-10, or ‘human assistance’ camera combines a camera featuring a 1/3-inch colour CMOS sensor, image recognition, and control into a compact unit to help to prevent errors in manual assembly operations.

A set of work instructions are loaded into the camera and displayed on a monitor which can be connected directly to the unit while the camera scans the work piece. No PC is required. The operator follows the on-screen assembly instructions and after every action the system compares the result to the correct stored image before the next step can be taken. If a step is incomplete or if a mistake is made, it is displayed to the operator so that it can be corrected. If everything looks OK, the operator can proceed to the next instruction.

Each step completed is verified and recorded in a CSV file including the time spent and an optional image of the inspection can be saved, as a digital audit trail. This means provides data that can be used for assembly work analysis and traceability. The system is also able to accept and check string data, so a keyboard or barcode reader can be attached to allow input of data such as part number, serial number and user ID. This data can also be logged as part of the digital audit trail, so it can simply be used just to log the serial number of the part being assembled. The unit is available in two magnification options, capable of covering most work assembly benches.

Image processing

The camera features in-built image processing software for pattern matching, colour verification or identifying the presence of textures. The pattern recognition tool is one of the most important tools and covers the majority of the applications for the Ricoh SC-10.

Examples include:

  • Ensuring a connector has been positioned in the correct orientation.
  • Checking the correct label has been affixed to a part.
  • Packing applications where the presence of a certain number of parts needs to be present within a box.

When the pattern recognition is coupled with the sequence checking process, the Ricoh SC-10 becomes a very powerful tool in ensuring assemblies are carried out correctly.

Examples of this include:

  • The presence and correct orientation of a part that may be hidden or obscured in later steps of the assembly.
  • Ensuring a component’s stability by checking the screws have been used and tightened in the right order.

The colour recognition tool is useful for assemblies that have parts with unique colours. Fuse box assemblies are a very good example of this as, although a lot are the same shape and size, their colours denote their amperage rating.

Improving car audio PCB assembly

The particular PCB contains a number of different components such as relays, condensers, and coils which require more than 50 visual checks to make sure that all the parts are present and correctly located. This high number of checks is demanding for a human operator and as a result, mistakes were often made, allowing incomplete boards to progress to the next stage of manufacture.

The introduction of the Ricoh SC-10 has made a major difference. The high magnification version was used for this application, making it easier to inspect the small components. In addition to checking the correct assembly of the components, it is even possible to check the status of jumpers and dip-switches. The system performs almost 50 checks in around five seconds, so the operational efficiency has also been improved.

Added benefits

In addition to avoiding errors in manufacturing, the system can help operators learn new tasks, since they follow the instructions step by step and can see if they do something wrong. The work instructions are set up for the particular task and are loaded into the camera either locally from an SD card or from a network via the inbuilt Ethernet port. This easy loading of instruction files means that they can also be created in various languages to accommodate multi-national workforces.