a few months ago

Your AI application in the real world. Here’s how.

Let’s talk about how you can bring your AI application into the world.

Machine Learning and AI projects are getting a lot of attention these days. A great deal of libraries are in the game. You can easily play around and fast enough create your on model, maybe even test it on some dataset on kaggle. But have you ever thought about taking AI to the real world, like self-driving cars?


If you’re about to dive into a machine learning or AI project, or better, you’ve already finished one, don’t you want this into the real world ?

How could we bring AI into the world?

A lot of libraries were created to do machine learning and deep learning tasks, you can easily create Neural Networks and train them on your data. Besides, you could also explore the world of Kaggle. But let's assume you're already have all set up. 

Within the python world, we have the most common libraries such as Keras, PyTorch, Scikit-learn and Tensorflow. Python is great, it has an awesome community built around it and a lot of books to play and learn.

But to bring the application we need to build a communication between the AI world and the Real world, one of the ways we could do just that is with serial communication. After that, the world is your oyster, you could set up a server and create a service, you could wire up a camera and set an alarm when you've detected someone with a CNN ( Convolutional Neural Network ). Besides you could also create a self driving car model, like this awesome example with NN.

Mobile devices

A great deal of fun is to provide this service also on mobile devices, they're growing quite commun on today's society. You could create your app and publish it on playstore. Imagine your friends using your AI application on their phones? How awesome would that be? 

But can we make this communication with python?

Yes we can! Python has a great library to make this communication, you can look the documentation here. We can add serial communication to the AI application and make it communicate if a lot of devices such as ESPs (used a lot in IoT applications) or Arduinos ( A great hardware plataform to start if you’re new in electronics ).

Low cost and Lots of fun

The great thing about those options is that they’re extremely low cost and pretty straightforward to start with even if you never played with hardware.

Ok, but how could we use this?

Let’s suppose we’re creating a robotic arm, if you haven't already check out the awesome robotic arm from OpenAI, for placing parts into cars in production system. We could use a camera and CNNs to detect if someone is in the workspace, this could set up an alarm. Without adding a great deal of sensors and a lot of wires across the room, with just a camera and a creative engineer. If you want more content like this, I have a YouTube Channel

It's great to train models

But it's even cooler to see our creations into the real world helping people with a better and secure life.