Question 1

Walton is thinking of improving its product inspection system. So far they have been manually checking every product with their human employees. One of their engineers pointed out that they can apply machine vision for their product inspection system and it can also be used for product measurement, counting, and determining the proper location of different parts of the products to make the whole system more efficient. The finance team disagrees by saying that it will cost more to buy these expensive machines and if the machines fail to properly detect the product and different parts of the product, then this will just be a huge loss for the entire company. The engineer replies that if they use a machine learning method with a huge amount of data labeling each product, then the machine will be much more accurate and the company will benefit enormously in the long run. The CEO is in a dilemma.

a) Which machine learning method is the engineer talking about? Explain logically. (Marks: 3)

b) What are the differences between the proposed machine learning method and unsupervised learning? Explain with examples. (Marks: 4)

c) How can machine vision make the entire system more efficient? Explain with different applications of machine vision. (Marks: 4)

d) Do you think a machine vision-based product inspection system is feasible for Walton? Briefly explain how a machine vision-based product inspection system might work in your own words. (Marks: 4)