Machine vision applications -  Successful examples from many industries


Successful examples from many industries

Vision system and robotics ensure 100% quality

February 2007

Johnson Controls uses 8 cameras to test up to 70 characteristics of car cockpits in 2 variants on the basis of 4 criteria. A VMT imaging system delivers 100% reliability.

Most vehicle manufacturers get the parts for their vehicle interiors from suppliers. This allows them to improve quality and reduce costs. Johnson Controls is one of the world's leading companies in the fields of automotive interiors, electronics and batteries, and the company counts major car manufacturers among its most important customers.

At its production facility in Lüneburg the company manufactures a great variety of dashboards and door trim panels for different automotive manufacturers. Johnson Controls supplies these parts directly to the manufacturer's production line. This means, of course, that they must be tested before they leave the factory in Lüneburg to establish that they are totally fault-free.

In order to keep one step ahead of its customers' growing expectations, Johnson Controls constantly strives to innovate in order to improve product quality and quality assurance.

Safety and quality requirements are becoming increasingly complex, and this applies particularly to dashboards. As a result, a very large number of quality-relevant parts that are important for subsequent production steps have to be tested before they can be delivered. These include correctly fitted rivets, particularly in the airbag area, clips, spring holders, other add-on parts and assemblies and screw fittings generally.

60 to 70 changing characteristics are currently tested. The fact that there are both left- and right-hand-drive models means this number is higher than it would otherwise be. The tests establish not just that the parts are present or fitted correctly but also that the correct parts have been used and that they are positioned accurately. Up to now, most of the characteristics have been tested manually by Johnson Controls staff working in three shifts.

The nature of the task means a mechanical solution is not feasible

To automate the testing of the dashboards, a flexible and extensible system had to be created that would ensure reliable and continuous 100% testing of the dashboards and allow a large number of tests to be carried out and documented on all types of dashboards.

A further challenge was posed by the different colour combinations and surfaces that have to be handled by the system. It was not possible to specify the type and colour in advance. The system also has to allow the faulty dashboards to be transferred to a reworking point and to indicate to operatives which faults exist, and where, so that they can identify and repair or replace the parts or assemblies found to be faulty as easily as possible and absolutely reliably.

Due to the high number of characteristics, some of which are difficult to access, and to the need for a high level of flexibility to allow tests to be added quickly and simply, a mechanical, tactile solution was not considered.

This demanding challenge was met by Weinheim-based VMT Vision Machine Technic GmbH in close collaboration with Dieffenbacher Automation GmbH, a company located in Seevetal near Hamburg. The project was handled by the engineers at VMT's north German office in Garbsen. The fact that the two companies were geographically close to the customer worked in the project's favour, and consequently the system was successfully implemented within a short space of time.

In order to come up to the high requirements stipulated, it was decided to put in place an automated system using a multi-camera version of the proven VMT IS image processing system, with image processing components from STEMMER IMAGING GmbH, in conjunction with an ABB robot. The system is completed by two conveyors. The first of these passes the good dashboards on to the next stage in the production process, and the second takes faulty dashboards to the reworking point.

"We get all the image processing components, such as the cameras, lenses and frame grabber cards, from STEMMER IMAGING. In addition to our own proven software tools, it was particularly helpful to be able to use the Common Vision Blox software platform," reports VMT chief executive Harald Mikeska. STEMMER IMAGING GmbH, which is based in Puchheim near Munich, supplies all the components required for image processing and develops the Common Vision Blox image processing software platform and associated tools.

The offline teaching function, in particular, has accelerated the optimisation of the system. This is made possible by the fact that all tests are recorded and the associated images saved.

Automatic operation starts

The new work sequence begins when the operative places the dashboards at the starting point. The robot then takes the dashboard and presents it in five different positions to a total of eight cameras fitted to a steel structure in the cubicle. It would not be possible to carry out all 65 tests within the very short cycle time available without so many cameras.

In the first step the dashboard type is identified by the image processing system, and the correct test for this type is started automatically. All the characteristics are then tested in a total of five robot positions. On reaching one of the positions, the robot starts a measurement via the Profibus interface. After only a few hundredths of a second it receives the signal to continue. In each position the VMT image processing system records and evaluates multiple characteristics simultaneously. The characteristics themselves are located on all sides of the dashboard. After carrying out the tests in the last position, the image processing system transfers all the measurements to the robot control unit. It takes approximately 8 seconds to evaluate the entire dashboard, and that includes all the robot's measurements.

On the basis of all the measurements taken, the dashboard is transferred to the reworking point conveyor if any faults are found. At the reworking point the operative sees the positions of the faults on a "digital" dashboard displayed on a large screen. Once reworking is completed, the dashboard is tested again. The image processing system remains available for automatic measurements throughout the reworking process.

According to Mr. Winkenwerder, Johnson Controls' project manager, the image processing system is so easy to use that his staff was able to create new characteristics and optimise the system during production after only two days' training. "The offline teaching function, in particular, has accelerated the optimisation of the system. This is made possible by the fact that all tests are recorded and the associated images saved," continues Winkenwerder. "Since going into operation the image processing system has proved to be highly reliable," says Winkenwerder, who is thoroughly satisfied with the solution, particularly in view of the fact that the expected cost savings have been realised.

The VMT image processing system

The real centrepiece of the system is the VMT IS software, which has been developed together with customers from the automotive and automotive supply industry over several years in over 500 projects. The development of a simple and intuitive user interface was a very high priority, and this enables the user to carry out tests independently after only a few days.

The system is set up and operated using the graphical user interface, without the need for any programming at all. You work with the graphical user interface in exactly the same way, regardless of whether you are using the system for robot vision guidance, completeness checking or plain-text reading, for example. The user languages (German and English as standard) can be switched over at any time online, and other languages can be added.


Kopenhagen, Denmark

For more than 50 years JAI has delivered industrial CCD and CMOS cameras with innovative engineering, high-end quality, and operational reliability and durability.

Ricoh Imaging

Hamburg, Germany

Since 1919 the Asahi Optical Co. has been producing optical components. In 1951 production of CCTV lenses was started, firstly under the brand name Cosmicar and later as Pentax.


Puchheim, Germany

STEMMER IMAGING has been leading the machine vision market since 1987. It is Europe's largest technology provider in this field. In 1997 STEMMER IMAGING presented Common Vision Blox (CVB), a powerful programming library for fast and reliable development and implementation of vision solutions, which has been deployed successfully throughout the world in more than 40,000 imaging applications in various industries.

CVB Minos - High-speed decision tree object recognition and OCR/OCV
  • High-speed object recognition software
  • Character recognition (OCR/OCV) under difficult conditions
  • Contains Minos Teach for learning new objects
  • Contains Minos Search, a tunable search engine
Ricoh lenses for 2 megapixel cameras - High resolution and compact design
  • Small, high-resolution, compact lenses
  • High detail resolution with enhanced contrast
  • Available with focal lengths between 12mm and 75mm
  • Locking focusing and aperture
CVB Image Manager
  • Platform for the development of image processing applications
  • Open programming environment
  • Simple to operate, combining flexibility with high performance