CHAI – CHai Made With AI
In every person’s life there is something is running in their back-up mind that is logic and technology. Hence, Technology is all about the developing Logic. Let’s talk about a cup of tea. Drinking tea makes us feel fresh and happy but making tea is also an art in itself. The ingredients and their wonderful aromas make up for a great experience. But to make this a much greater experience, here I introduce you the CHAI made by a robot and created with all the AI capabilities. Just imagine, when the artificial intelligence will decide what tea is to be made and when it will make the perfect tea for you and your friends by taking different orders and serving each one of them with their desired tea, then you will surely feel proud to drink and to serve CHAI to your friends.
So let’s begin creating this magic. The CHAI with AI, can be made with the following three main steps:
* CHAI AI Functionality
* AI Software program
* AI hardware - Creating a Robot
Out of the above three, the main ingredient is the ‘CHAI making functionality’. Once this is clear, then our most of the work is done. Later comes how to code that in a specific software, and then feed that in a Robot, that will make the tea. But in this article we will only look in to the first step and see how end to end functionality is to be created, so that a Robot can understand and make our required tea.
The very first time when I heard the word “computer” and “robots” that has fascinated me since then, was in 1986 in an Indian Bollywood movie, where the actors say, its not 1986 but a computer age and that what all you do, you must do it in a robotic style. Those days the robots used to speak in an echo double coarse voice and their walk used to be “slow and break” style. Those days I used to feel that if I must communicate with the robots, I need to change my style of talking and walking, to suite Robotic behaviour.
But recently I saw a television program where they introduced very young, school/college going students who have done some innovations. These talented students had built their own robots performing some useful functions. The household things done by their robots or some automatic task like remotely starting of lights and fans. All this show how AI and robotics has entered in everyone’s life today, and what all new things are been created.
I remember in my school-boy days, we were told to write a composition on “A Robot in My House”. With lot of enthusiasm amongst all of us, that we will be writing something on Robots, the word that was very new at that time, we all started writing and putting all our imaginative and creative ideas for our Robots. The next day at school when our English teacher asked if we had written our essays, and to her surprise, for the first time everyone had done their homework. When our teacher read few essays, she was disappointed. She asked us, who is the robot in your house who does all this work. I was expecting at least one of you to have understood that and write it. When everyone went blank, she told us, that it’s your mother. We all were feeling embarrassed that day, that we could not understand that much. Yes, our mother is the robot for all us, doing multitasking and playing multi- roles. But speaking about the automation and robotics, we can see its advancement and progress in our day to day life now.
My mother is very fond of tea. Just recently I have started preparing the morning tea. Many a times it so happens that I make very nice tea with all the ingredients properly put in, and properly boiled, but sometimes it also happens that the tea that I prepare is not at all good. The ingredients are not put in proper amounts, or it’s not boiled for enough time finally ending up in a bad tea. In doing this, I told my mother, better I will make a robot who will make tea every time with proper ingredients that you decide. Then there will be no complaints about this. My mother was surprised and was wondering if that is possible. I told her, that if we take help of AI and robotic technology effectively then we can really make this possible but that will surely take lot of efforts to be put it.
Working on Functionality:
From the above thing, I got an idea to create a “RoboTea “, the robot making tea. I started thinking about how we write a program for this and what all things will be required to do so. Data Science, AI, Machine Learning and Robotic technology all put together will surely make this possible. But in this article, we will put aside Robot making and will only focus on the software and the functionality to prepare the same.
Artificial Intelligence can give us the solution to above problem. The following are the few technologies we will be using to solve this:
* Data Science
* Machine Learning
The above two solutions are for functionally tackling the problem. The we can program our understanding using Python , which is a very suitable software for writing our Artificial Intelligence functions.
The Science dealing in data, from its creation to its manipulation to its retrieval for the solutions, all together constitute of Data Science. Since the advent of computers, Data was everything and now with data science in picture, data has become more important than anything else in our world. Hence any data, even if it is not required, should not be deleted. I remember in my entire career I never deleted any data from production, or even acceptance environment. For Insert and update of data also I would properly check and then execute those steps. If in case we need to delete any data in any environment we should be 100% sure about the reason and the function. I have seen very rarely that the project demands any data deletion from any environment. That is precisely the reason why when we delete anything, we get a pop up window asking, “are you sure you want to delete this”. Make sure in your data science projects you do not delete any data. Data is fuel , data is everything.
Data Science in a Nutshell
For the above example we will develop a small data, to show, that when you put in the different measures of Sugar, then we will get the following different types of sweetness’s in our tea. (see Result column) . This is called “Features-Label data set”, to help AI system to decide the type of tea sweetness it requires. Here we have 4 features: Water, Milk, Sugar and Tea, and the Label is the Result. In our example we have only changed Sugar feature, and others are kept as same, but if we change other features, then we will get different types of teas, that we don’t look in this example.
Now once these figures are put in the Robot, and then if we order our Robot to make a sweet tea, it will go through its menu and check, and this will itself decide what measurements of all the ingredients should be considered and will finally make the desired tea.
The above example is a supervised Machine learning, and the algorithm used is Classification. The above example teaches the machine, that if you need to make a sweet tea, then you require to put 1.5 spoons of Sugar, not less not more. It is clearly demarcated and classified in to different values that will generate different results. The above is called Training data.
Once the machine gets to know this, it will then start understanding, if a particular command is given then what all is to be done related to it. Vice versa, if following ingredients are added then what will be its result? That is called Teaching data. So, for example if I now add this data.
1.7 spoons of sugar, then it should understand from its data that it should be sweet tea. But when I add 1.8 spoons of sugar, then it will be very sweet tea, and if the person who is ordered sweet tea, this tea will be discarded, and a new tea will be made. This way it will make that the tea prepared is always perfect.
Conclusion: This is the simplest example of the AI and machine learning. In actual world, the data that the data scientists use is huge, it has thousands of columns and millions of rows, forming a very huge data set for a specific problem. I have only showed you how these algorithm work with each other to make a machine understand and perform its tasks correctly. There remains no error in giving the solution, since the learning keeps happening daily, every time the tea is ordered, or made, the data is added. Just imagine the sort of excellence and perfection that machine will gain daily from the data input and data output, and finally it will be completely error free.