Androtics is a firm to create all kind of projects related with robotics, artificial intelligence (AI) or neuroscience.
The main goal is to make real any idea in mind! For instance, the creation of different robots from scratch or the implementation of different state-of-the-art AI algorithms. All done with 3D printing, Arduino, open source libraries and ROS. The main goal is to make real any idea in mind!
Hi! My name is Andres and I am the founder of Androtics. I am spanish and I live currently in Bonn (Germany).
One of the greatest questions with no answer is how intelligence works and emerges in any of us; this is something I’m sure everybody ever asked. The popular saying «you don’t really understand something until you do it by yourself» reflects why the artificial intelligence field is growing more and more nowadays.
Androtics emerged from my pasion to robotics, which has been one of my hobbie for the last years. Robotics shouldn’t be motivated just to create intelligent machines, but also to understand ourselves.
As many other stories, mine starts after finishing high school. I considered my self (and I still do) a very curious person that always tries to understand everything and how it works. That’s why I decided to study a bachelor in physics in the Complutense University of Madrid (UCM, Spain). Thanks to physics, I could understand a little bit more the basic principles that govern our universe at big and small scales, doing my final thesis about general relativity in the University of Calgary (UofC, Canada), where I studied the 4th year of my bachelor.
During the bachelor, I remember one day that I tried to program a chatbot. A chatbot is a computer program that you can «have» a converstation more or less coherent with, using the keyboard. I still didn’t know anything about artificial intelligence, but I was really excited with the idea of having a computer answering to my questions (maybe I was too bored). My problem was that I didn’t know how to program a computer, so I started to learn all the necessary to develop a simple chatbot in Matlab (a programming language). This moment was an inflexion point for me to the computational world. I started to show great interest about programing and automation, discovering the arduino and raspberry pi platforms, which can be used to control motors and read sensors in a very easy way. These platforms uncovered me a new world of possibilities! and I thought that I could build a physical body to my chatbot in order to build a real robot. I found out that Matlab wasn’t the most proper programming language to do it, so I begin to learn Python y Linux. After some months of hard work, I end up with a talking robot that was able to move with help of some motors, although it was very limited due to my little knowledge in robotics.
Just after finishing my bachelor in Physics, I was selected in a summer program in the Helmholtz-Zentrum Berlin (HZB, Berlin) to work in the development of an interface to control a photon detector in the synchrotron BESSY II. During 8 months, I had to learn low-level C++ programming from scratch, which has been really useful in the following years. During this period I also had to decided what to study next: either a master in theoretical physics or a master related to robotics. Finally I decided the last choice studying the Master in Inteligencia Artificial in the University Politècnica of Catalunya (UPC, Barcelona).
For one year and a half that lasted the master, I learned the base of the current algorithms in artificial intelligence. Something to emphasise was my final master thesis (TFM) that I realized in the Institut de Robòtica i Informàtica industrial (IRI, Barcelona).
This project consisted in develop a robotic system that helped people with limited mobility in putting on hospital crocs shoes, using a WAM robot shown in the image above. The robotic system was able to understand some voice commands and to track the body parts of a person in real time. Everything was integrated in ROS (Robot Operating System), that I learned to use at the same time of the development following the tutorials in the webpage. This project received an award to the best master thesis 2017 in Cataluña and was calificated with honors.
After finishing the master in February of 2017, I decided to create Delta, which was the first robotic project with pretension to state it in internet. The idea was to create a short video where I show all the building process and a small demostration of its funtionalities. The video was published in the Youtube platform where I created the «Androtics» channel. Since I received good reviews, I decided to create more content in the channel and convert Androtics in my firm, not only to share my projects in robotics but also to motivate people to do their own projects. For this reason all projects in Androtics are open source and anybody is able to do them at home.
Finally, to understand the following step of my career, I need to give my own opinion about the current situation of artificial intelligence, stressing that this is just my opinion and conclusion obtained along all my experience:
Artificial intelligence have done great and very useful advances in the last decades. But, artificial intelligence nowadays is only applicable to very specific optimisation problems, i.e., in finding the most optimal solution to a very concise problem. Artificial intelligence is very useful to create a mathematical model that extracts features and patterns in big databases, and to describe correlations within the samples of this data. For better understanding, let’s suppose that we create a neural network (mathematical model) to recognize people (extraction of features and patterns) in different images (database). This problem is very specific: now if we want to recognize dogs with the same neural network, our model will fail. In addition, we don’t know exactly what is going on in this neural network during the people recognition: we don’t know how information is encoded and how the network «looks for» persons in the images, simply it does. We don’t know either how to integrate and interpret information coming from different sources, for instance, how to create a neural network that integrates sound from a microphone (auditive source) and images from a camera (visual source) in order to recognize people moving and talking. On the other hand, in biological systems such as animals, the the previous situation is happening all the time, for instance, an animal that sees and hears another animal can think that the second animal is a predator so it runs away to save its life. We find this kind of intelligence in any animal, and the impression I have is that artificial intelligence is little or nothing motivated by biological intelligence. In my opinion, in order to create something as complex as biological intelligence, we need first to study and understand it to finally reproduce it.
Said this, I decided to focus my career in computational neuroscience, where I obtained my current PhD position in the research group Neural Circuits (Caesar, Alemania). My goal and motivation is to understand how biological intelligence emerges from neurons and I sincerely think that this field will contribute a lot to artificial intelligence in the future, and also it will help to understand and treat brain-related diseases. Besides this, robotics is still a great pasion for me, and I love to learn and to keep up to date in this field. This is what motivates me to make Androtics grow.