We are buzzed with the two words in our everyday life. One is Automation and the other is AI. These are not the unfamiliar words which are discovered or invented in recent time or we are just now implementing. These are words which are in the existence for long time. Right from an invention of wheel till invention of internet technology we are driving the automation. Now, the question comes why do we want to attain the automation and what makes the AI.
I'm publishing this
article as part 1 to discuss why Automation is related to AI and which area we
should be focusing first to make sure what is the right way to achieve the
right thing. Nothing is conclusive, and therefore, this article is also not
conclusive. All I'm trying to put on this article is my understanding and the
way I understand and achieve my goals of Automation and AI.
To understand all
these, let's start understanding what is Automation and AI.
As we understand the
meaning of Automation, let us look at different meanings in multiple perspective:
Automation is…
"where human
intervention is less."
"where machines
work continuously without input from the operator."
"where action is
performed in the workflow with the input of the completion of one
process."
"where
controlling is done without the execution of the sets of instruction."
"where reaction
of one process is autonomous in nature and does not impact the reaction of
another process."
"where operation
is completed with less amount of time."
….and we can have
more definition on automation.
What we have
understood from the above multiple definition is, automation is active
mechanism to do the specific task.
Now, let us talk
about the definition of AI. We all know AI is Artificial Intelligence.
Therefore, why the intelligence is artificial. Let's check different definition
of AI.
AI is…
"which can solve
the problem."
"which can learn
from the behavior."
"which can mimic
human behavior."
"which can
interact with other systems based on its need to perform its tasks."
"an intelligent
agent which can exhibits its characteristics to interact with other intelligent
agents."
…. a lot more….
When we read the
above definition of AI, what we can observe that there are some kind of
instructions which is driving the AI. If we move these instructions into one
block and run them based on time, we are automating these instructions and
making these automation tasks as driving force for an AI.
We can derive
multiple types of automation and through those multiple types of automation, we
can have multiple type of AI created for our tasks or products. Now, what are
these multiple types of Automation. When we search on this in the internet we
get all these….
·
ANN
- Artificial neural network
·
DCS
- Distributed Control Systems
·
HMI
- Human Machine Interface
·
SCADA
- Supervisory Control and Data Acquisition
·
PLC
- Programmable Logic Controller
·
Instrumentation
·
Motion
Control
·
Robotics
·
Factory
Automation
What are all the
above automation being about? These are all set of instruction which gets executed
within time. Now my question is, are all these can be considered as AI or all
the above type of automation can be the driver for AI. To answer this, now, let
us understand the type of AI……
·
Weak
AI - non-sentient machine intelligent, typically focused on a narrow task
·
Strong
AI - sentient machine with consciousness and mind.
When we talk about
Weak AI, we think about the inline instruction which are limited to do the
specific tasks or to complete one cycle of process. Then the automation scripts
or programs which intend to perform small tasks or becomes a part of the
complete process leading to become the driver for weak AI. These automations
could be continuing w.r.t time.
Now, when we think
about Strong AI, we think about the instructions with the power to relate to
other set of instruction and then they can even derive its own instructions to
complete the entire process of completing the tasks. These automated scripts
can be self-developed to make themselves more intelligent and it is repetitive.
With the above we can
say that any automation can lead to specific type AI and thus become the driver
for AI.
Considering the type
of Automation and AI, different companies like Microsoft, Amazon, Google,
Facebook and many more has come up with their frameworks to integrate with
different software applications and hardware. These frameworks have made the
entire process of adopting AI in our application software and hardware simple.
In my next article,
I'll be more focused on the implementation by taking one of the technology and
see how we can build the AI to achieve specific tasks.
