Accelerating Performance with the Artificial Intelligence of Things
traffic conditions and other quality-
of-life factors.
In semiconductor manufacturing of
wafers and dies, many tests cannot
be run until the packaging phase.
With computer vision, it is practical
to take images of the wafer earlier in
the production process and inspect it
for issues. The inspection can be used
to find defects and determine the
number and location. This will allow
an earlier determination of final yield
from the wafer. With previous
labeled images of diagnosed defects,
it is also possible to classify the
defect types using computer vision.
This will help augment a larger root
cause
analysis
for
process
improvement.
Figure 2: Silicon Wafer and Wafer Defects
production quality. For example, in
aerospace and automotive parts production,
you can determine incomplete finishes or
casting issues.
Defect detection in discrete parts – With
visual inspection of discrete parts, you can
more easily catch several production
defects. These are various issues in
IIC Journal of Innovation
- 38 -