Artificial intelligence and cognitive robots – insides from the expert.

Expert interview with Milad, Head of Artificial Intelligence

(Simplified responses understandable for everyone)

What exactly are your tasks at NEURA Robotics?

I run the artificial intelligence (AI) department at NEURA Robotics and have been with the company since day one, along with our CEO, David Reger, and around twelve other people. Part of my job is to tackle novel and ambitious goals and turn them into viable projects. These ideas can come from David himself, other employees, or the AI team members. For instance, when making a robot cognitive, imagine you want to have a conversation with a robot and have the robot understand and respond to you, then definitely AI is necessary. Based on this task, I create a series of projects that in a step-by-step manner enables us to realize different versions of our final goal. My team and I carry out various projects with the overall goal in mind. Of course, these projects are difficult to implement, but we put in our best to make these projects possible and generalize it. A typical situation is when we build a robot to recognize and pick up a glass, we generalize it so that it can recognize and pick up anything else, a plate, a bottle, a teddy, anything. This is part of the process we accomplish when making robots cognitive.

How have your experiences been since the first day at NEURA Robotics?

The first day I spoke with David and heard his super ambitious goals I was shocked. This is largely because I was the one who was going to implement a substantial portion of these goals together with my team. I honestly could not sleep well for a while. It was the first time in the world that a company decided to try something like this, and there were no guidelines or best practices available. What I was most concerned about was how we could execute our mission with the right people. But David kept us motivated, which was extremely helpful. Yes, we had some bad days and many exceptionally good days, but in the end, we tried not to repeat any unpleasant experience and learn as much as we could from it. What is particularly important is having lots of flexibility in thinking. Being flexible as an expert and relying on yours and others’ experiences makes everything possible.

How would you define artificial intelligence?

Artificial intelligence is easily understood with the word cognition. Cognition might be a little vague, but if you compare artificial intelligence to natural intelligence, you understand that natural intelligence is so natural to us humans that we do not think about it. Artificial intelligence is inspired by the human brain. We imitate the human brain mathematically to figure out how it works and transfer that knowledge to machines in a step-by-step process. Neural networks are used to develop an uncomplicated way to replicate neurons mathematically. Recently, new hardware development such as graphics processing units (GPU) have been used to develop deep neural networks with many neurons that work better than engineering systems. Artificial intelligence means that a machine can perceive, understand, listen, see its environment, plan its next action, and execute it.

Why do we need artificial intelligence?

We need artificial intelligence because there are many problems which sophisticated technology cannot solve. For instance, in physics, if you want to design a car that moves fast, it can be designed with the equations of physics so that the result is precise. This though will be only an approximation. But in robotics, we need to compensate for that approximation, and the only way to do that is with machine learning. With AI, you do not have to find an equation, you can create a huge data set, pass it to a d and the model understands it in ways you cannot even explain. For example, if you want to recognize a cat in a picture without AI, you will have to say that in your code: a cat has this type of eyes, this color, and it looks like this. But if another cat comes tomorrow, this method cannot recognize it because it is smaller and has a completely distinct color. With AI, you can create a dataset of many cat images, train a simple deep learning model with it, and it will recognize all the cats. It is like the natural intelligence of humans. As a child you see lots of cats or cat pictures, and every time you ask your mother what it is, she says it is a cat, but she does not explain to you what a cat looks like. After a while, you can recognize any type of cat.

Can you give examples of some real-world applications?

AI can help in many industry sectors. Especially in applications that are typically repetitive and require humans to conduct the same repetitive task. Firstly, humans do not like repetitive tasks because they get boring very soon, and secondly, often these tasks are even difficult and dangerous for us humans. We should know that recently robotics is considered as a part of big AI topic. Robots are not classically AI enabled or cognitive but here I am talking about robots with integrated AI in it. Cognitive Robots can be used for all kinds of repetitive and difficult tasks or even to rescue people in dangerous situations like fires or after floods and earthquakes. These types of tasks are what humans have always wanted to automate. Smart multi-purpose robots could take care of the elderly too. During the Covid 19 pandemic, robots could have aided in testing people and in hospitals to help the patients.

Is artificial intelligence related to machine learning in robotics? And how?

Machine learning is a subset of artificial intelligence. AI is broader, but when scientists talk about machine learning and deep learning, they are referring to specific aspects of AI. With machine learning and deep learning, we develop algorithms and mathematical toolsets that are predictive and trainable and, for example, can be used to build cognitive robots.

Can you give insights to some accomplishments that you and your team have achieved for our cognitive robots?

We have developed a lot in a brief period over the last three years. For example, by using a 3D camera, we can recognize many objects in front of the robot or easily train new objects too. The robot can also distinguish, select, and pick up any objects selectively. Our robots can selectively choose a specific object, such as glass instead of tape or a water bottle. We can train a completely customized grasp only by showing it to the robot without needing to write a single line of code. This simply means, not only can the robot grab any object, but even from a specific angle or point. For example, if there is a glass cup on a table, the robot must know how to grasp it and how to place it properly on the table. We are constantly developing and improving everything we have achieved. Additionally, we can run all our AI tools offline because the methods are all integrated and optimized inside the robot controller. Running everything locally on the robot is safe and ensures that the data on the robot cannot be accessed through the Internet. Imagine if you could run a computer offline – that is exactly what we made possible. Our robots have voice commanding and gesture control and can respond to users or even tell a joke. I would like to reemphasize that I am not going into the technical details here, rather I focused on bringing examples that are understandable to everyone.

ball balancing

What are the components of our robots that give them cognitive abilities?

The first part that makes the robot cognitive is the integrated sensors. Robots should be able to sense their surroundings, see the environment and respond when spoken to. The 3D camera, the microphone, and specific, combined with technology we developed ourselves, like touchless safe human detection, are a must have for a cognitive robot. The optimized processing software and hardware are used to understand the data, and the sensors are constantly collecting data, learning, and adapting.

What are your plans for the future of robots at NEURA Robotics?

Cloud solutions are a fantastic opportunity for our robots, and we are researching and developing in this direction. We are not waiting for others to come up with a solution in the future, we are designing our future now. Furthermore, we are also working on other projects that I cannot mention, but another thing we are working on is continuous learning techniques, which we are improving so that our robot is constantly learning from human actions. That is really a big step. We are working on a new generation of service robots that can be used in offices or our homes.

How do you test the intelligence of robots? Is there a way to measure their intelligence?

It can be practical, for example, if we want to see that the robot understands us, we can talk to the robot, and it will answer. Additionally, we have a testing software in which we have stored a lot of data using machine learning. This data set is divided into a training set and a test or evaluation set. With these sets, you test whether what you did is correct or not. There are at least two types of tests: qualitative and quantitative. Qualitative is the practical form of testing, while the latter is quantitative, which means you can run a lot of test programs, but they must be done correctly. We have many test codes that are usually developed with any software module.

In your expert opinion, what is the future of the robotics industry with artificial intelligence?

The future holds huge possibilities. The exciting but challenging thing is the fact that we can have all cognitive capabilities in our robots. The goal is to help humans with repetitive, boring, or dangerous tasks. Robots that will help human workers in all fields, just as we see phones everywhere today. This is the simplest but easy-to-understand example because I am sure everyone remembers the time when there were no phones or only a few non-smart phones, but now we cannot live without them. That is why I see a future where we cannot live without intelligent robots because they will make our lives easier. They can help us at home, in factories, hospitals, nursing homes, schools and everywhere else. To operate and interact with robots, people do not need any skill except normal talking, showing, or explaining. We are developing robots that are versatile and can do everything. That is my vision for the robotics industry.

Read more about NEURAs technologies HERE.

Milad and robot