Enterprise Europe Network

Computer vision-based solution sought for quality control of metal stamped parts

Country of origin:
Country: 
SPAIN
Opportunity:
External Id: 
TRES20190725001
Published
26/07/2019
Last update
31/07/2019
Expiration date
31/07/2020

Keywords

Partner keyword: 
Process automation
Optical Technology related to measurements
Other scanning related (incl. image processing, ...)
Industrial measurement and sensing equipment
Process control equipment and systems
Machine vision software and systems
Forging, pressing, stamping and roll-forming of metal; powder metallurgy
EXPRESS YOUR INTEREST

Summary

Summary: 
A Spanish (Catalan) company specialist in the design and manufacture of thermal, thermo-acoustic and electromagnetic insulations for automotive industry, is looking for a computer vision related technology for quality control in metal stamped products. They are interested in either a technical or a research cooperation agreement.

Description

Description: 

The Catalan enterprise, founded in 1984, is a global Tier 1 leader providing shielding solutions to reduce heat and noise and eliminate electromagnetic interferences by means of metal based stamping parts. It currently manufactures more than 22 million shields per year, and it is a worldwide player with strong presence in the European, Asian and North American markets. They own the following certifications:

• IATF 16949 - Quality management system for automotive production
• ISO 14001 - Environmental Management Standard
• OHSAS 18001 - Occupational Health and Safety Management

The company has an automatic cold stamping line, which produces more than 15 different references. An operator placed at the end of the automated production process (figure-01), grabs the stamped parts that leave the machine on a conveyor belt and visually inspects them one by one to detect the most common defects, namely:

• Cracks of 0.3 mm in length or greater with different shapes.
• Overlappings.
• Lack of material.

Now, the company wants to improve the quality control implementing a computer vision-based technology to detect defects, which is mainly expected to:

• Perform an objective and constant unitary verification of the parts.
• Satisfy the increasingly demanding quality requirements of their market.
• Reduce the ocular fatigue of the operators.
• Allow savings in materials and energy consumption by early detection of the root cause of defective parts.

The technology sought should be fixed on the machine soundproof enclosure (figure-01), and be able to continuously monitor the parts moving on the conveyor belt to detect defective units.

The company is interested in a research or technical cooperation agreement with a partner with whom to collaborate in the development and implementation of the computer vision based solution.

Technical Specification or Expertise Sought

Cooperation sought: 
The company is seeking a computer vision solution able to detect defects such as cracks, overlappings or lack of material in metal stamped parts, which: • Are made of aluminium or stainless steel. • Have dimensions (width x height x depth) from 300mmx200mmx150mm to 1500mmx500mmx250mm. • Have an irregular shape with depressions and mounds (figure-02). • Are stamped in a process with a cycle time of less than 3 seconds per part. • Leave the stamping machine impregnated with stamping oil, on a conveyor belt that vibrates slightly. • Are obtained in a line that works three shifts a day. Therefore, light conditions may vary during the day inside the production plant. Technologies that have been tested with unsatisfactory results: • Thermography: the resolution is low. • Laser: the reflection on the surface of the parts causes errors. • Infrared (IR): the large diversity of references to be checked, with different geometries and positions on the conveyor belt, makes it difficult for the IR backlight to transmit through the defects in either a controlled and uncontrolled environment. • Red, Green, Blue (RGB): the error margin measured in both controlled and uncontrolled environments, is higher than acceptable limit of 5%. • 3D Vision: the percentage of error during scanning on the conveyor belt is too large.

Partner sought

Cooperation area: 
Start-ups, SMEs, universities or R&D institutions partners with deep knowledge and experience in computer vision technologies, are sought to jointly develop a solution under an agreement that can range from research to technical cooperation, depending upon the stage of development of the provided technology.

Type and size

Cooperation task: 
SME 11-50,University,R&D Institution,SME <10,SME 51-250