How VisionPro ViDi by Cognex performs Defect Detection using Artificial IntelligenceEstimated Reading Time: just 4 min

Artificial Intelligence (AI) is the intelligence demonstrated by machines, in contrast to the natural intelligence displayed by humans and other living beings.

For a machine acting like a human means being able to perform 5 macro-functions:

  • Perception (recognizing and interpreting reality)
  • Reasoning (drawing new conclusions)
  • Planning (detecting patterns and adapting to new environments)
  • Knowledge Representation (extract and store knowledge)
  • Natural Language Processing (communicate successfully in a human language)

AI research started in 1956 after a workshop at Dartmouth College and its fathers where:

  • Allen Newell and Herbert Simon from CMU
  • John McCarthy and Marvin Minsky from MIT
  • Arthur Samuel from IBM

Dartmouth College group focused on Restricted AI that solves only one of the 5 acting like a human macro-functions at a time.

Recently, a huge breakthrough came with the birth of Machine Learning.

Machine Learning is a field of artificial intelligence that uses statistical techniques to give computer systems the ability to learn from data, without being explicitly programmed.

This step ahead allowed systems to solve problems that involve:

  • Supervised Learning
  • Clustering
  • Dimensionality Reduction
  • Structured Prediction
  • Anomaly Detection

Another breakthrough came with Deep Learning.

Deep Learning is a subset of Machine Learning that, using architectures based on Neural Networks, allows machines to learn a large variety of activities as humans do.

This step ahead allowed systems to solve problems that involve:

  • Supervised/Semi-Supervised/Unsupervised Learning
  • Image/Audio Recognition
  • Natural Language Processing

Nowadays, there are 3 main problems that can be solved with AI

  1. Regression Problems: this could be applied whenever the output of a subset of inputs is a continuous function
  2. Classification Problems: this could be applied whenever the output of a subset of inputs is a discrete function
  3. Clustering Problems: this could be applied whenever the output of a subset of inputs is not predictable/known

The 3 problems are solved by using sequences of algorithms called Neural Networks.

Usually, every Neural Network performs 4 activities:

  1. Data Collection
  2. Data Cleaning
  3. Patterns Identification
  4. Prediction Making

One of the most interesting applications of Artificial Intelligence is how Cognex performs defect detection.

Cognex (Cognition Experts) is a manufacturer of machine vision systems, software and sensors based in Natick, Massachusetts.

It was founded in 1981 by a lecturer in human visual perception at MIT (Robert J. Shillman) and two of his students (Bill Silver and Marilyn Matz).

The company’s first product was called Dataman and was released in 1982.

It was an Optical Character Recognition (OCR) system designed to read, verify, and assure the quality of letters, numbers, and symbols printed on products and components.

Through the years, the Cognex Corporation grew through developing its products that can fall into 3 main categories:

  • Vision Systems (Vision Sensors + Vision Softwares)
  • Barcode Readers
  • 3D Laser Profilers

The company grew through the years becoming in 2017 a company with

  • More than 1 thousand employees
  • More than US$ 700 million of Revenues

Counting on a diversified accounts base in several industries such as:

  • Automotive
  • Consumer Electronics
  • Consumer Products
  • Electronics Products
  • Electric Vehicle Batteries
  • Food and Beverage
  • Life Sciences
  • Logistics
  • Pharmaceutical/Medical
  • Product Security Solutions
  • Semiconductor
  • Solar

In the Vision Systems product category, Cognex developed VisionPro ViDi, a deep learning-based image analysis software.

VisionPro ViDi is meant to use Artificial Intelligence in order to perform:

  • Defect detection
  • Texture and material classification
  • Assembly verification and deformed part location
  • Character reading

This could be done with 4 different functions:

  • Locate (to find and to count objects and features)
  • Analyse (to find differences and defects versus the normal appearance of an object)
  • Classify (to build classes of object, i.e. to separate stains, hits and scratches)
  • Read (to decipher codes)

The target is to transform previous activities from highly supervised to partially supervised ones.

VisionPro ViDi works in a 2 stage process:

Step1. Training Phase

Step 2. Deployment Phase

The Training Phase exploits the Deep Learning algorithms to teach the machine how to perform a specific task.

Deep Learning algorithms are able to teach something to a machine by using statistical techniques that give computer systems the ability to learn from data, without being explicitly programmed.

Required steps are:

  1. Load some hundreds of sample images to be studied
  2. Characterize the type of function to be performed (Locate, Analyse, Classify or Read)
  3. Train the machine by analysing manually some sample images
  4. Letting the machine to deliver results based on what learned in point 3, on a subset of the loaded images in point 1
  5. Manually validate the outcome of point 4
  6. Letting the machine to evaluate another subset of images and re-start with point 5

Training phase can be performed in:

  • Supervised Mode (every outcome of point 4 is evaluated manually in point 5)
  • Semi-supervised Mode (only the outcome of point 4 for which the machine is not sure are evaluated in point 5)

This phase takes some hours in order to be effective.

The Deployment Phase is the integration of the Training Phase into the day-by-day process.

Required steps are:

  1. Acquire Images with a consistent process (i.e. the light has to be always the same)
  2. Let the machine Analysing and Interpreting the Images
  3. The machine delivers a result of its analysis with a related degree of certainty to allow an operator to intervene and manually examine some spot cases

The Deep Learning algorithms will keep learning so that the machine gets more and more reliable over time.

As per every Vision System using AI, VisionPro ViDi has super-human capacity because of:

  • Ability to acquire images with quality and detail that is far greater than human eyes
  • Operates consistently 24/7

The key point for everything to work, though, is to set up an image acquisition process that is extremely consistent over time. Failing this point will make the machine unable to process the input data effectively.

Nicola Zaffonato Administrator

Business Strategy | Product Marketing | Executive Master eCommerce Management | Business Innovation Master | MSc

I am driven by my personal growth and of people/contexts that surround me.

I followed a professional path in Valentino Fashion Group and Luxottica during which, thanks to the ability to understand different businesses and interests, I was able to succeed in Operations, Merchandising and Retail.

These organizations have exploited my ability to mediate and translate needs/constraints into practice, assigning me to Project Management roles.

Luxottica relied on my ability to analyze, to anticipate things and to imagine/implement solutions by appointing me in Supply Chain Management department and assigning me to the Product Management of IoT solutions for Anti-counterfeiting and Retail digitalization.

During this professional path, I also developed my leadership by managing teams to build Processes, Organizations, Systems and Governance Tools.

follow me
Sharing is Caring!
search previous next tag category expand menu location phone mail time cart zoom edit close