By 2022, the number of operational industrial robots worldwide should exceed the Berlin population. Nowadays, in our factories, flexible cobots work hand in hand with people and automatic learning helps them become even more versatile. It is very fascinating to follow the developments in the automation industry at a rapid rate.
When I visited the first laboratory From Toshio Yanagida to the University of Osaka in 2010, I was very surprised by the way the researchers managed to handle unique proteins using laser traps. I was almost also surprised that a very cute robot and inspired by the man calmly waited in their conference room to greet me. However, this fortuitous meeting, with the only exception of my parents’ robotic lawn mower, was the full extent of my practical experience with robots until recently.
In the meantime, however, I must admit that I could have developed a serious crush on the subject of robotics. While the use of robots escapes from mass production using specialized machines for low -volume manufacturing, an interesting trend emerges – robots begin to become more flexible and sometimes they have a kind of integrated automatic learning to perform simple tasks. Slowly, but the machines are surely more intelligent. Naturally, as a professional (ex-) experienced automatic learning, I am more than enthusiastic about the promise that resides in the intersection of robotics and AI. But before discovering the interesting world of artificially intelligent robots, let us first build a base and understand the ways of the industrial robot industry.
Robots have become an integral part of manufacturing since the first robotic company opened its door in the 1950s. They are one of the main reasons for the ever -increasing quality and productivity that we have known in recent decades. In 2022 the International Robotics Federation (IFR) expects an amazing number of around 3.8 million operational industrial robots worldwide.
Fig. 1: Annual shipping of industrial robots (source MCKINSEY))
Who can locate a trend in Figure 1? The growth of annual shipping for industrial robots is starting to speed up a considerable speed after 2010. IFR statistics Show that this is to a large extent due to technological catch -up in China where the number of robots increases at an impressive rate of 30% per year. There, already 27% of all operational industrial robots were found in 2018.
Fig. 2: Global market for industrial robots: key companies by report returned 2017 (source Statist))
The biggest, but by far not the only Manufacturers of industrial robots are Abb,, Omron,, Fanuc,, Kawasaki Robotics,, Kuka and Yaskawa. In Figure 2, we can see the revenues of these robotics giants. Their main customers are always the highly automated electronic, automotive and metallic industries.
In recent years, Chinese producers have required their piece of cake. Although their market share is still small, manufacturers of industrial robots like Honyen, Siasun or Estun slowly wins market share on their domestic market. To have a feeling of proportion, however, allow me to mention that in the first three quarters of 2019, Fanuc alone has sold more than 3.5 times The number of robots in China that all of its Chinese competitors. As industrial robots have a lifespan of decades, long -term proven reliability and the robustness of the holders always seem very convincing.
There are different types of robots such as Scara,, delta Or Cartesian coordinated robots. An intuitive example is a supposedly articulated robot (see fig. 3). It is modeled after a human arm and with its six degrees of freedom, this industrial robot is very flexible and therefore among the most popular robots for industrial applications. Robots like this are used for welding, assembly, sealing, handling of materials, picking, cutting, paint or spraying. We have probably all seen this type of robots in videos of the automated automated production lines of car companies or similar advanced industries.
The design of the robot consists of five main components: controller, sensors, manipulator or robot arm, final effector and drive.
Fig. 3: Example of an articulated robot
Controller
The controller can be considered the robot brain. It can be connected to a computer and serves as an interface between the person configuring the system and the robot itself.
Sensors
The robot perceives its environment via sensors. It can be microphones, cameras or pressure sensors. Reception of sensors can be used by an industrial robot to perform its scheduled task.
Manipulator / robot arm
A robotic arm is used to position its final effector. In our example of the articulated robot, the arm is modeled by the human arm and imitates the shoulder, the elbow and the wrist. He has six degrees of freedom and therefore has a wide range of possible movements.
Final effector
The final effector acts like the robot hand and is located at the end of the arm. Depending on the robot’s objective, the effector can be a welding torch or a pliers.
To drive
The robotic player is the engine that moves the coins from the robot. The common discs are powered by hydraulics, electrically or pneumatically.
By reading this article, you can think: “Well! Where can I buy a robot to automate my tasks?”. But be aware! This industry is very advanced and complicated.
In an ideal world, you could buy an ABB controller, sensors and a Yaskawa robot arm, your Kuka’s favorite final effector and a Siasun journey and combine your eclectic robots to meet your very specific needs. In the real world, however, most companies use different programming languages and all their APIs are not open. Manufacturers of robots want to be more than just producers of robotic parts and aim to push the software side of their business. This is why in a production chain, you mainly see complete robots from different suppliers working next to each other. Fortunately, there are attempts like the robot operating system (Ros), which, despite its name, is rather a middleware than an operating system, aims to unify the level of the software. However, it is not clear at the moment how its adoption of the market will develop.
Companies that keep a cool head and help you set up your production chain are systems integrators. They will come to your establishment, will understand your needs and make the equipment of different suppliers together work together. Their business is to prepare your production chain and they are paid like kings. If you are interested in the biggest systems integrators and their income, do not hesitate to consult this List by control engineering.
There are different use cases and automated manufacturing strategies. They all have strengths and weaknesses and involve very different costs and deadlines.
Special use machines
Personalized machines have been the catalysts of mass production and, in many cases, it is always logical that factories design and build special machines. A good example is a factory for cigarettes that almost certainly uses specialized cigarette machines. Product ranges in these factories change approximately once every two decades and high initial investments are recovered in the long term. Due to specialized machines, very few workers are necessary to produce large amounts of cigarettes. Such tailor -made robotic systems make sense for broadband systems with a long service life.
Traditional robotic systems
A slightly different case is the production of cars where a new model can be produced every two years. In this case, the production lines must be adapted to each new model. It is not economical to design and produce custom machines for each part of each new car. Automobile companies therefore make more flexible uses of robots such as our articulated robot represented above. This flexibility, however, is delivered with a price. Because these robots can be reused, the configuration of a new mounting chain is complex. System integrators can invest months to set up and program robots in a complex production line. Initial high investment must be amortized over the years. Such a use case is characterized by standardized parts and production life cycles of the order of several years. Traditional robotics is designed to operate independently with the security ensured by the isolation of human contact.
High and low volume manufacturing
While mass production is an obvious case of robotic automation, the scenarios currently high and at low volume are mainly carried out manually. In this area, the initial investment of programming and implementation of robots in the traditional way may not bear fruit. This is where we see a lot of development: the conviviality of robotic software is considerably simplified and technologies such as automatic learning are used to teach the tasks of robots that are not clearly defined. An emerging trend is the so -called cobots – collaborative robots. Currently, they represent 3% of the robotic marketgrowing slowly. Cobots are intended to interact with humans and work near them. A company that aligns this long robotics tail is the Danish Universal robots. For example, their UR5E is shown in Figure 4. Isn’t it beauty?
Fig. 4: COBOT Ur5E By Universal Robots
Even scratching the surface of the subject of industrial robots, it already becomes clear that robotics is a fascinating area. Interesting things are necessarily to happen if robots become more versatile and intelligent. I am particularly delighted to see how flexible cobots will be used and adopted in smaller production lines and how robots will find their way in our daily life.
In follow -up articles, I will plunge a little more into the competitive landscape, the global market and the startups that seem to change robotics for the best.