I.

Introduction to robotics and automation

Robotics is a field of science working with machines that perform tasks based on predetermined and adaptive programs and algorithms in an automatic or semi-automatic way. These machines – commonly called robots – are either controlled by humans or work entirely under the supervision of a computer application and algorithms. Robotics is a comprehensive concept that includes the building, planning and programming of robots. These robots are in direct contact with the physical world – and they have often been used to perform monotonous and repetitive tasks instead of human beings. Robots can be categorised based on their size, application domain or purpose, and we will discuss this later.

Robotics vs automation

Automation is a much broader concept than robotics. It means that specific parts of a process or an entire process is performed without human intervention. Instead, the process is operated only by predefined or adaptive computer applications and electrical or mechanical machines. The predefined applications refer to algorithms, in which all the operations are predefined and executed independently regardless of any unforeseen changes in the environment. Adaptive automation means that the algorithm can change its behaviour according to changes in the process or environment.

Robotics goes hand-in-hand with automation, as in most cases robots are part of an automated system. Although there are times when robots are used with little or even no automation – and you can also have automation without robotics – the two are like twins that have a lot in common, but each with their own distinct personality.

Note

If there are machines that perform tasks for us, why do we even work? The fact is that the capabilities of robots are limited. Even if robots look smart, these machines are typically only good at some very narrow application domain. Even if we apply an armada of robots for the many tasks in our life, these domains do not overlap in a way that makes a complete system that can completely replace humans.

In other words, as these robots can only operate in a very limited domain – and we are nowhere close to achieving general machine intelligence – you should not be afraid of evil robots or robot domination.

The types of robots

Robots can be classified in different ways. We’ll look at four main methods of categorisation:

  • Size

  • Application domain

  • Purpose

  • Number

Size

When looking at size, the following categories exist:

  • Nanorobots or nanobots: nanorobots are made of nanomaterials and range in size from 0.1 to 10 micrometres (to get an idea of how small that is, a human red blood cell is about 5-10 micrometres). Nanobots are in the early research stages – mostly, the concept is being discussed for use in medicine and many more years of hard work are needed to make them a possible solution. One vision for nanorobots is to inject them into a patient’s body to identify and cure diseases.

  • Microbots, millibots and minibots: these robots are very small but still larger than nanobots – and they actually already exist. Microbots, millibots and minibots are smaller than 1 mm, 1 cm and 10 cm, respectively. The smallest flying robot is RoboBee, with a wingspan of 1.2 cm and a weight of 80 milligrams. The wings can flap 120 times per second and the robot can be controlled remotely. The goal of such a small device is to form a flying swarm for search and rescue, or artificial pollination.

RoboBee, the smallest flying robot
RoboBee, the smallest flying robot

  • Small and mid-sized robots: these robots are typically under 100 cm (small) or about the size of a human (100–200 cm, mid-sized). Most household robots, toys and social robots, humanoids (robots that have a similar appearance to humans – the Transformers of comic books and films being a common example), and digital personal assistants are this size. Small and mid-size robots are the kind we see and meet most of the time – in movies and in real life.

  • Large robots: these robots are bigger than us. Much bigger. There are some large humanoid robots, even up to 8–10 meters. However, humanoid large robots are typically made for research purposes, or just for fun. In fact, most large robots don't look like humans – they are made for automation in manufacturing, construction, agriculture, autonomous driving and navigation.

Application domain

It is also possible to categorise robots according to their application domain, dividing them into personal and industrial robots.

  • Personal robots are used in our daily life and are designed to be useful for individual or family use. Non-technical people can operate personal robots to perform repetitive and perhaps boring tasks to save time or to entertain us. Household robots, social robots, digital personal assistants and toys are the most common personal robots.

  • Industrial robots are robust and are created to perform specific tasks in a pre-programmed manner in manufacturing, construction or agriculture, for instance. Applications include assembly, disassembly, mounting, screw tightening, welding, painting, visual inspection, and so on. Industrial robots are outstanding at one specific task: these are fast, precise and reliable machines. Without industrial robots, we wouldn't really have today's level of technological development.

Purpose

Another possible categorisation for robots is purpose. Robots can have a specific and a general purpose. But what does that mean?

  • Task-specific robots: these machines perform a particular task or a sequence of possible tasks. It can be as simple as a robot arm that moves objects from A to B, but it can be as complex as a social robot with an advanced natural language interface. The construction and behaviour of these robots cannot be changed; they follow predefined programs according to their original purpose. Household robots and industrial robots are among such machines.

  • General-purpose robots: in this case, the task of the robot is not predefined. Various components of the robots can be bought separately, and these components can be assembled in different ways in order to solve specific tasks. The components may include robot arms, wheels, cameras, step motors and additional sensors and actuators. These robots may also have wireless connections, such as Wi-Fi and Bluetooth. The "brain" of the robot – which is generally a small computer – can be “trained” to perform different tasks with different components using custom applications written in computer programming languages. Common programmable small computers – also called embedded systems – are the Nvidia Jetson and Jetson Nano, Raspberry Pi, and Arduino. These embedded systems have general-purpose input and output connections (GPIOs) to which sensors and actuators can be connected using a standard communication interface.

Other general-purpose robots include a prebuilt body comprising sensors (like cameras and microphones) and actuators (like arms and legs). By developing different computer applications, the robot can perform different specific tasks. Examples of such robots include Softbank Robotics’ Nao, Pepper and Romeo, or Boston Dynamics’ robot ‘dog’, named Spot.

Number

Robots can also be categorised according to how many there are:

  • Single robots: a single robot works on its own. It has a duty which it performs based on a predefined program. The predefined program might involve advanced technologies which make it able to adapt to its environment, and the robot might be connected to the internet, but the robot is still alone. Even if there are several single robots in one place, they are still ’alone’ as they cannot communicate with each other.

  • Robots in teams: robots can work in teams, just like humans. Often a task is done in sequence by several robots. Think about video recordings of how cars are assembled. The chassis is welded, then comes the doors, then the car is painted, front and rear windows are next, and so on. All of these steps are performed by different robots that can only do that particular task.

  • Swarm robotics: robots can also work in a swarm. In this case, a large number of simple robots are controlled collaboratively. Individual robots in the swarm are not particularly valuable, but the swarm itself can perform important tasks. Just think about bees in nature. A single bee cannot do much, but without millions of bees working in swarms, humans probably wouldn’t even exist. Possible applications of swarm robotics are exploration and rescue, microbiology, surveillance and pollination. However, at the time of writing (2021) swarm robotics is mostly in the research phase.

The evolution of robots

The word robot comes from the Czech word "robota", which means “serf work” in Czech. Karel Čapek’s 1920 play, where machines take over the world, made the word “robot” widely known. But humanity has always been interested in the rethinking of human existence. Even before the 20th century there were several attempts to recreate a human being and legends telling of people who had succeeded. One of the most famous ideas belongs to the 16th century alchemist Paracelsus. He stated that a small human-like being (called a homunculus) could be created in a flask using only chemical procedures. Later in the 16th century, the word golem entered public consciousness. Based on a folk story, the golem was made of clay and would serve people if someone inserted a special parchment into its mouth or forehead. The story says that after a while, the golem confronted its creator and eventually turned against him.

Looking at the history of robotics, there is a universal interest in imbuing robots with humanity or some human attributes. This interest generally has three main conditions:

  • the robot has to be similar to a human being in some way (in appearance, in thinking, etc.)

  • the robot has to be better at something (stronger, smarter, etc.)

  • the robot has to be completely under the control of its creator

There was a milestone in the history of robotics when machines that were stronger than people appeared. Machines that replaced a human's contribution to work appeared during the first industrial revolution around 1769. At that time the main purpose was to reduce costs and the time spent on production and to increase the quantity of products without human interaction. Automation became the main concept at that time. With automation, several processes can be completed without any human intervention. As work was done by machines, it led people to find new ways of working and living. Machines do not get tired like people do, so machines can work 24/7. The risk of error and the amount of waste also decreased with automation.

Robots are also characterised by controlled accuracy and effectiveness. In the 1800s, computer technology was not present. However, people were able to create large machines to perform complex tasks. After 1950, there has been an important development in robotics.

Example
Shakey the Robot, first general-purpose mobile robot
Shakey the Robot, first general-purpose mobile robot

Shakey

The first industrial robot "Unimate" was created and used in automobile factories as a replacement for the manual workforce, and the first general-purpose mobile robot "Shakey" was also created. Shakey had built-in cameras and touch sensors and was able to interact with its environment.

Various positive statements can be made about the existence of robots, but humanity is not 100% satisfied. The labour market is under continuous pressure from people willing to work, as workers doing repetitive tasks could be replaced with machines. A certain fear always appears in connection with robots about them replacing the human workforce, or if robots may have more control over humans than desired.

As robots get more realistic, another fear arises. People generally tolerate robot-like robots. Our brain can easily categorise robot-like humanoid robots like we categorise industrial robots in manufacturing. People may get confused and even frustrated when meeting an overly realistic robot. In this case, we know that it's a robot. However, the brain can't really deal with this fact due to its realistic appearance. Its skin, movement and even speech is very similar to that of a person, but our brain struggles with the categorisation: is it really a robot? Does it have thoughts? Can or should I trust it?

Note

These fears could be a reason for creating the "rules of the robots" in the middle of the 20th century, also known as The Three Laws or Asimov's Laws, created and named after the author Isaac Asimov.

The three laws are:

  • A robot may not injure a human being or, through inaction, allow a human being to come to harm.

  • A robot must obey the orders given it by human beings except where such orders would conflict with the First Law.

  • A robot must protect its own existence as long as such protection does not conflict with the First or Second Law.

(Asimov 1950)

Another milestone in the history of robotics was when the first remote-controlled mobile robot discovered the moon's surface around 1970. A bit later, in 1986, Honda started a project to create humanoid robots that look similar to people. The evolution continued, and robots appeared in more and more fields like healthcare, manufacturing and logistics. The evolution of robots is still an ongoing process and now robots are present in our daily lives. Robots are in homes (vacuum cleaners), in workplaces (assembly robots) and in healthcare (social robots in patient treatment or surgical robots), for instance.

Humanity is in the fourth industrial revolution, which integrates the hottest emerging technologies, like robotics, IoT, 5G, artificial intelligence, and many more, to take the industry to new levels.

Next section
II. How do robots work?