What are automated vision systems? What capabilities do they have and should you consider investing in one?
Automated vision systems, also known as machine vision systems, provide image-based inspection for a variety of industrial and manufacturing applications. Machine vision systems typically involve a smart camera that has been calibrated to perform specific inspections automatically as products flow down a production line. 2D and 3D cameras are now commonly used for automated inspection, along with robot guidance, product sorting, quality control and much more.
Types of Automation Vision Systems
There are thousands of applications for vision systems in manufacturing. The type of vision system most appropriate for your application will depend on several factors:
- Type of Inspection (i.e. label inspection, code inspection, defect inspection, etc.)
- The specific properties of your process or products (size, shape, orientation, etc.)
- Line Speed (speed of inspection)
- Budget
- Other timing, production, or monetary constraints
Common types of inspections are listed below. These are achieved with either 1D, 2D, 3D or area/line scan cameras.
- Print & Code Inspection Systems: Identify incorrect or damaged labels on products before they are released. Examples include: Detecting damaged, poor quality text; Checking 1D barcodes and 2D date matrix codes for readability and product match; Inspecting preprinted artwork, ink-jet quality, laser markings, chemical etchings, etc.; Verifying correct labeling, nutritional information, correct ingredients, etc.
- Defect inspection systems: Catch and resolve product, packaging, or surface defects in real time. Examples may include: dent inspections, label placement verification, broken parts, broken seals, surface deformities, missing parts, etc.
- Custom Packaging Inspection: Limit the false reject rate and meet the exact quality and standards needed for your product. Examples include: Inspecting for dents, defects, damage, holes and creases for plastics, metals or glass; Checking labels for wrinkles, position, skew, date codes; Examining molded features, missing caps, correct seals, crooked caps, etc.
- Custom Web Inspections: Eliminate tears, holes, improper webbing or missing/defective features in web materials. Examples include: Real time image processing to detect holes, tears or damage to web, etc.; Inspection of artwork, barcodes and 2D codes, print inspection; Real-time process feedback to remove damaged sections of webbing, etc.
- Process Control Vision Systems: Utilize process feedback to achieve a flawless process line with an industrial vision process controls. Examples include: Precise control of robot positioning with imaging systems; Capturing and communicating measured values to various data logging systems; Historical data capture from your process; Data trends logging for process monitoring and proactive maintenance; Event monitoring and event capture systems, etc.
- Custom Tolerance Measurement: An integrated part inspection vision system ensures product quality, repeatable production standards, and can reveal valuable process information. Examples include: Verifying part tolerance, thread count, machine vision measurement and statistics; Checking final machined part tolerance; Comparing finished products to CAD images; Camera & laser scanning; Infrared cameras and radiometric analysis, heat seal inspection, etc.
While these are just a few of the common categories of vision systems, we see custom applications every day. Capabilities for vision inspections continue to evolve every year, and provide unique, efficient solutions to manufacturing or quality problems.
Examples of Automated Visions Systems in Manufacturing
Below are two examples of how an automated vision system works in an industrial environment:
Label Inspection & Barcode Evaluation for Children’s Consumable Product 
Barcode verification for correct product labeling was needed on a high speed juice bottling line for a children’s consumable product. A smart camera was integrated to read the 2D bar code on each bottle as it passed by. This system verified the label was correct and present, and matched the product being filled into the bottle.
This system prevented product waste from incorrect labeling and ensure consumers got the right product with correct expiration dates and ingredients at the store.
Crushed Bottle Identification System
The client needed to eliminate crushed bottles on the manufacturing line, preventing them from being packed into cases or shipped to customers. Various sized bottles in translucent, frosted and/or made from dark, opaque colors passed through the line daily. Translucent objects are difficult for camera lenses to “see” and crushed bottles do not take a predictable shape and the high speed of the line made it a difficult challenge for human staff.
This challenge was resolved with a custom bottle inspection vision system capable of handling clear, frosted and opaque bottles of varying sizes with a budget friendly solution. Every bottle passes in front of a backlight, creating a contrasted view of the bottle. The camera takes an individual snap shot of every bottle and processes the images inside the camera Bottle images not within the normal range of shapes triggers a bottle reject that pushes that bottle off the line before the filling station.
Solutions Provided by Automated Vision Systems
As demonstrated above, automated vision systems provide image-based inspection for a variety of industrial and manufacturing applications. These solutions can range from standard print inspections to creative visual inspections of challenging product issues. We invite clients to experiment with new processes, technologies and techniques that push the efficiency limits of traditional quality control systems by contacting an EPIC Vision Expert today.