Industrial robots

from Wikipedia, the free encyclopedia

An industrial robot ( IR , also: industrial manipulator ) is a universal, programmable machine for handling, assembling or processing workpieces . These robots are designed for use in an industrial environment (e.g. automobile production ). They belong to the mechanical engineering discipline of automation technology . The industrial robot generally consists of the manipulator (robot arm), the controller and an effector (tool, gripper, etc.). Often robots are also equipped with different sensors . Once programmed, the machine is able to carry out a work process autonomously or to vary the execution of the task within limits depending on sensor information.


ZIM 10 welding robot from VEB Robotron-Elektronik Riesa in the GDR , 1983

The origin of industrial robots can be found in reactor technology, where hand-operated manipulators were used early on for tasks within radioactively endangered rooms ( hot cells ). In this context, Raymond Goertz constructed a teleoperator arm in 1951, which allowed operations on z. B. to carry out radioactive material.

George Devol

The industrial robot was officially invented in 1954 by George Devol , who applied for a patent for a programmable manipulator in the USA. Together with Joseph F. Engelberger , Devol founded the world's first robotics company Unimation in 1956 . The company developed the Unimate industrial robot , which was first used in a production line at General Motors in 1961 for removing and separating injection molded parts. The first commercially available robot was introduced by Planet Corporation in 1959. This robot was already suitable for simple tasks such as resistance spot welding. However, Planet Corporation's concept was still based on mechanical control using cams and limit switches, while Unimate already had numerical control.

The first industrial robots in the automotive industry were equipped with hydraulic cylinders as drive sources. Hydraulic industrial robots were used in Japan from 1967 and in Germany at Mercedes-Benz in automobile production from 1970. In the mid-seventies, electric actuators with microprocessor control prevailed, which are still used almost exclusively today.

In 1973, the German robotics pioneer KUKA built the world's first industrial robot with six electromechanically driven axes, known as the Famulus . A year later, in 1974, the Swedish ASEA (now ABB ) presented their also fully electrically powered robot (IRb6).


Industrial robots are supplied in different designs and by different manufacturers. They are usually purchased as a standardized basic device and adapted to their respective task with application-specific tools.

A distinction is made between robots based on the kinematics used :

An important parameter of industrial robots is the payload . This describes the maximum mass that can be attached to the end of the manipulator. Articulated arm robots currently range from 2.5 to 1300 kilograms. In addition, the dynamics and the accuracy are decisive factors.

The collaborative robot is a special type that is designed in such a way that it can work together with people in a room without a protective device. This opens up completely new application possibilities, but also brings with it new requirements for the safety concept, which can lead to restrictions in terms of load capacity, cycle time, etc.

application areas

Articulated arm robot with welding gun (2004)

Industrial robots are used in many areas of production, such as B.

  • as a painting robot for painting or as a robot for polishing
  • as a measuring robot for measuring and testing
  • as a grinding robot for belt grinding

Structure and structure

Structure of an IR

The structure of an industrial robot (IR) includes:

  • Control : It monitors and specifies the movement and actions of the IR. This requires programming .
  • Drives : The drive moves the links of the kinematic chain and consists of a motor, gear unit and control. It can be driven electrically, hydraulically or pneumatically.
  • Internal sensors : This provides information about the position of the kinematic chain. It is used by the control to compare the target and actual position. Internal sensors can be, for example, incremental rotary encoders, interference patterns or light barrier functions.
  • Kinematics : It represents the physical realization of the load-bearing structure and creates the spatial allocation between tool / workpiece and production facility. It consists of rotary and translational axes. As a rule, at least 3 degrees of freedom are required to be able to reach each point in space. This requires at least 3 axes of movement.
  • Gripping systems : A gripping system establishes the connection between the workpiece and the IR. This can be done via force pairing, shape pairing or material pairing.
  • External sensors: It gives the IR feedback about the environment. It enables a flexible reaction to unplanned changes. External sensors can be, for example, image processing systems (e.g. laser light section systems), triangulation sensors, light barrier functions and ultrasonic sensors.
Quick tool change system consisting of robot side, tool side and tool tray
  • Optional quick tool change systems: they enable a program-controlled tool change e.g. B. in welding, cutting, joining, palletizing, gluing. The quick-change systems, which are usually of modular construction, consist of at least one robot side, several tool sides and a corresponding number of tool trays. Depending on the area of ​​application, the tool changer can be equipped with media couplings (water, hydraulics, air), electrical signal connectors (fiber optic, data bus ) and electrical power connectors.


The manipulator or robotic arm is a multifunctional handling machine that consists of a series of rigid links that are connected to one another by swivel or sliding joints, whereby the joints can be adjusted by controlled drives. One end of this "link chain" is the base, while the other end is freely movable and is equipped with a tool or gripper for carrying out production work.


For the creation of robot programs , there are the online programming and offline programming methods , which are often used in combination.

Most modern robot controls contain a complex programming environment into which other tools can be integrated. As a rule, there is a seamless integration of modules by means of which external sensors such as. z. B. a camera system or a force / torque measuring system, can be integrated into the robot program. This is how the behavior of the robot can adapt to external influences:

Example: Components can be recognized by sensors such as a camera and processed individually by the robot. “Pick-and-Place” (= take and place) robots can correctly fill a candy box from a mixed bulk material consisting of different pralines.

Example: Different screws can be fastened by means of a torque transmitter depending on their size.

The robot controller is often connected to a PLC . This regulates the interaction between the robot and the surrounding system technology.

Robot programming languages

Manufacturer programming language Control surface on the touch panel
Omron V +
Fanuc Karel
Kuka KRL
Yaskawa Motoman In shape
Stäubli VAL3
Universal Robots (UR) script Polyscope
Epson SPEL +
Denso Pac (RC7) & PacScript (RC8)
Mitsubishi MELFA-Basic

Online programming

The robot is programmed directly on or with the robot itself.

The online programming methods include:

  • Teach-in procedure
  • Playback procedure
  • manual input via buttons and switches (outdated)

Teach-in procedure

With the teach-in process (short: teaching), the programmer moves the robot to the desired position using a control panel. All coordinates (points) reached in this way are saved in the control. This step is repeated until the entire work cycle has been run through.

Playback procedure

The programmer follows the intended path by guiding the robot arm directly. The robot repeats exactly these movements. This method is often used with painting robots.

Playback method with 3D measuring arms (mobile coordinate measuring systems KMG). While the CMM is being guided along the component contour, coordinates are recorded and later converted into a robot program.

Offline programming

This is not required to program the robot, the program development takes place offline on a computer that is independent of the robot, so the robot can continue to operate during development, there are no downtimes.

Offline programming methods include:

  • Textual programming
  • CAD-based programming
  • Macro programming
  • Acoustic programming

Textual programming

The tasks are described on the basis of problem-oriented language. The process is comparable to programming in a high-level programming language.

Advantages of textual programming
  • The program can be easily changed and well documented.
  • The program can be created without using the robot.
Disadvantages of textual programming
  • A qualified programmer is required for programming
  • Almost every manufacturer uses its own programming language.

CAD-based process

In the CAD -based programming the robot to a PC workstation (often also Unix - workstations ) based on design drawings and simulations programmed.

The entire sequence of movements is already defined on the PC in a three-dimensional screen environment. As a rule, the surroundings of the robot and its tools are also shown. This allows various examinations to be carried out:

  • Determination of the correct position of the component in space.
  • Is the workpiece or the tool being moved?
  • Are the desired working points even achievable?
  • How much time does the robot need for this sequence of movements?
  • Does this program have any collisions with the environment?
  • Review of alternatives, e.g. B. a "smaller", cheaper robot can be used for the originally planned model.
  • Ensuring the feasibility of the robot application.
Advantages of CAD-supported programming (in connection with simulation)
  • The robot can already be programmed if it has not yet been set up.
  • Planning and construction errors can be identified at an early stage. Necessary changes can still be made on the computer at this stage and there is no need for expensive modifications on the construction site
  • Extensive changes to robot programs are sometimes much easier to make than directly on the robot
  • In the 3D environment on the computer, every part of the robot environment can be viewed from all sides. In reality, the working points of the robot are often hidden or difficult to access.
Disadvantages of CAD-based programming
  • The exact environment often does not exist as a 3D model. Load-bearing elements of the factory installation such as pillars, girders, trusses, etc. are therefore difficult to take into account.
  • All devices and tools must correspond exactly to the models in the computer.
  • Flexible supply lines ( compressed air feed, cooling water hoses , welding power supply or glue feed, stud feed at Studwelding) can only be inadequately mapped on the computer, but they cause considerable movement restrictions in the real industrial robot.
  • The Teaching is often easier and faster.

The program created in CAD-supported programming is transferred to the industrial robot via data carrier or network and can then be executed immediately. Usually, however, various adjustments ( robot calibration ) are still required, since the simulated environment never exactly matches reality. The connection to the PLC is usually only "on site".

Macro programming

For frequently recurring work processes, macros are created that represent frequently used command sequences in an abbreviated form. The macro is programmed once and then inserted at the required points in the control program.

Acoustic programming

The program text is programmed using natural language with the aid of a microphone. The system can acoustically confirm the commands and thus enable the correct detection to be checked.

Benefits of acoustic programming

  • Avoidance of input errors
  • greater freedom of movement for the operator
  • Adaptation to the usual natural form of communication

Disadvantages of acoustic programming

  • relatively high error rate of today's speech recognition systems

Coordinate systems

The tool position of an industrial robot is about his so-called end-effector ( English Tool Center Point briefly described TCP). This is an imaginary reference point that is located at a suitable point on the tool. Due to the historical development of industrial robots, it was common to define an electrode of a spot welding gun as a TCP. The tool or tool coordinate system was thus tacitly defined. The origin is congruent with the TCP. The Z direction points to the other electrode. The X-direction is orthogonal to the surface formed by the electrode arms.

In order to describe which position the robot tool should assume, it is sufficient to define the position and orientation of the TCP in space.

The position of an industrial robot can only be described in relation to the axis .

For each individual (linear or rotary) axis of the robot, it is indicated in which position it is located. Together with the structural lengths of the links, this results in a clear position of the robot flange. This is the only way to describe the position or configuration of the robot's kinematic chain.
The TCP or the tool coordinate system can only be described in relation to space. The basis is the Cartesian coordinate system. Transformations are used to switch between the axis-specific description of the robot and the spatial description.

The programmer determines to which point in space the robot tool is to be moved and how it is aligned. The robot controller then uses the so-called Denavit-Hartenberg transformation to calculate which position the individual robot axes must assume. See also inverse kinematics , direct kinematics .

Various coordinate systems are available for the spatial description of the robot position, which the programmer can use as required. The name can vary depending on the robot controller:

Spatial robot coordinate systems

World coordinate system

The world coordinate system (WORLD) usually has its origin in the rotary center of the first axis (base axis, base frame). It is the main coordinate system, which is immutable in space. All others are related to this coordinate system.

Base coordinate system

The basic coordinate system (BASE) is mostly used on the workpiece or the workpiece holder in order to teach point coordinates in relation to the workpiece or the workpiece holder. In this way, the point of origin of the basic coordinate system can be shifted and the associated point coordinates move with it. In the default state, the base coordinate system is congruent with the world coordinate system (base x 0, y 0, z 0, a 0, b 0, c 0). Several of these coordinate systems can be created in the robot system and saved with names. When programming, you can switch between the various basic systems.

Tool coordinate system (tool)

The tool coordinate system is located on the tool of the robot. Its position is defined by the TCP (Tool Center Point), which is located at a suitable point on the tool. Its orientation is determined by the so-called direction of impact of the tool (+ Z) and a second freely selectable tool axis. The position and orientation of the tool coordinate system are defined as a translational and rotary shift to the center of the flange plate.

Because the tool coordinate system moves with the tool, it is always the same relative to it, even if its position in space is variable. With a sensible definition of the TCP, the programmer can rotate the tool around its operating point or perform linear movements precisely matching the tool position. This suits the human way of working and thus makes teaching easier.

Several tool coordinate systems can usually be created in a robot controller, which can be selected using a tool number. This makes it possible to work with several different tools (e.g. two differently shaped welding guns ). These can be attached to the robot at the same time, but this can lead to problems with weight and accessibility. Alternatively, a tool change system can be used in which the robot can dock and undock different tools.

The so-called “external tool” is a special application of the tool coordinate system. Here, the TCP is not defined on the robot tool, but on the working point of a stationary tool. The points of the robot program are not fixed in space, but “stick” to the workpiece moved by the robot and are moved with it to the stationary tool.

In this case, for example, the robot does not move the tongs to the sheet metal, but rather moves the sheet metal held in a gripper to the stationary tongs.

External coordinate system

Another variant are external coordinate systems. In this case, the coordinate system with the workpiece connected , which is mounted on a simple 1- to 4-axis manipulator. Such manipulators have loads of up to 60 tons. Alternatively, the workpiece can also be manipulated by one or more robots, which must then be in communication with the robot that moves the tool. With both variants, several robots can work simultaneously on one and the same workpiece.

Base coordinate system

The base coordinate system can be freely positioned in space by the programmer, for example parallel to a device at an angle. Position and orientation are independent of other coordinate systems, but mathematically relate to the world coordinate system. If a robot program is defined in the base coordinate system, it can be easily shifted and rotated in space by simply changing the position of the coordinate system, but without having to re-teach a single program point (see base coordinate system).

Axis configuration

The movement of the robot tool via inverse kinematics leads to some special features. While a certain position of the axes clearly results in a position of the tool, the position of the axes for a certain tool position is not always clear. The system is computationally ambiguous.

Depending on the position of the target point and the mechanics of the robot, there are often several axis configurations that lead to the desired tool position. It is up to the programmer to choose the most suitable configuration. The controller must then ensure that this configuration is retained for as long as possible during the movement. Otherwise, switching between two configurations can lead to a very large movement of the entire robot for a minimal tool movement. This unexpected movement costs time and is often not possible without collision.

In some robot kinematics (e.g. 6-axis articulated arm robot) there are spatial points that lead to so-called singularities . A singularity is characterized and recognizable by the fact that two axes of the robot are collinear (aligned). A typical configuration with a singularity is the upside down position of the tool. Axis 1 and axis 6 are aligned here. The control cannot clearly assign axis 1 or axis 6 to a rotation to be performed around the vertical. Another constellation is given at the zero crossing of axis 5. Axis 4 and axis 6 are aligned here. There are an infinite number of axis positions that lead to the same tool position or trajectories in which several axes would have to be moved against each other at infinite speed. Some controls abort the program when passing through such a point.


In the sense of robot programming, transformations are the transfer of the description (position and orientation) of an object from one reference system (coordinate system) to another. For example, the position and orientation of the tool represented in the world coordinate system can be transformed into a representation in the workpiece coordinate system. With the so-called forward or forward transformation, the position of the robot flange is transferred from the description using axis values ​​to the description in robot world coordinates. This transformation is clear .

The reverse or inverse transformation transfers the description in robot world coordinates into the description using axis values. This transformation is ambiguous . In the early days of robotics, the transformation equations were formed from sin and cos terms of the respective axis values. The Denavit-Hartenberg Convention describes a set of rules with which it is possible to describe robot kinematics in a generally valid manner using matrices. This fundamental work, as well as the development of computer technology, made it possible to formulate the forward transformation as a matrix multiplication . The inverse matrix is ​​required for the inverse transformation, which is known to be numerically unambiguously solvable only under certain conditions.


Personal safety is very important in robotics. As early as the 1950s, Isaac Asimov established three robot rules in his science fiction novels , which basically state that a person must not be harmed by a robot or its inactivity. Today it is laws (in Europe the Machinery Directive 2006/42 / EG formerly 98/37 / EG) and international standards (e.g. ISO EN 10218 formerly DIN EN 775) that define the safety standards of machines and thus also of robots .

The dangers emanating from robots consist in the complex movement patterns, which are often completely unpredictable for humans, and strong accelerations, with enormous forces at the same time. Working next to an unsecured industrial robot can quickly prove fatal.

The first protective measure is therefore usually the separation of the movement area of ​​humans and industrial robots by protective grids with secured protective doors or light barriers . Opening the protective door or interrupting the light barrier causes the robot to come to a standstill immediately. In special operating modes where humans have to enter the robot's danger area (e.g. when teaching ), an enabling button must be pressed to explicitly allow the robot to move. At the same time, the speeds of the robot must be limited to a safe level.

More recent developments ( assistance robots) go in the direction that the robot uses sensors to recognize an approach of a foreign object or a person in good time and slows down, stops or even retreats automatically. This means that in the future it will be possible to work together with the robot in its immediate vicinity.

All control circuits with functions for personal safety are usually designed and monitored redundantly , so that even a fault, for example a short circuit , does not lead to a loss of safety.

The hazards emanating from the robot or additional systems are determined via a hazard analysis and a suitable protective device is designed for this. All devices that are switched in the safety circuit must correspond to the selected category.

Market structure

In the 50 years from 1961 to 2011, a total of 2.3 million industrial robots were installed worldwide. The most successful year so far was 2011 with around 166,000 newly commissioned industrial robots, 28,000 of which were in Japan, the largest robot country, and 25,000 in second-placed South Korea. China, USA and Germany follow in 3rd to 5th place. The International Federation of Robotics estimates that fast-growing China will be the largest robot sales market by 2014 at the latest.

The world market leaders in 2010 were the two Japanese companies Fanuc and Yaskawa Electric (with the Motoman brand ), each with a share of around 20 percent, and the German manufacturer KUKA Roboter with a share of around 10 to 15 percent.

Number of newly installed industrial robots worldwide every year
year Asia Europe America Worldwide
1998 069,000
1999 079,000
2000 099,000
2001 078,000
2002 069,000
2003 081,000
2004 097,000
2005 120,000
2006 112,000
2007 114,000
2008 060,000 035,000 017,000 112,000
2009 030,000 020,000 009,000 059,000
2010 070,000 031,000 017,000 118,000
2011 089,000 044,000 026,000 159,000
2012 085,000 041,000 028,000 154,000
2013 099,000 043,000 030,000 172,000
2014 134,000 046,000 033,000 213,000
2015 161,000 050,000 038,000 249,000
2016 200,000 056,000 038,000 294,000
2017 280,000 067,000 046,000 313,000
2018 283,000 076,000 055,000 414,000


Well-known manufacturers of industrial robots are:

Almost every manufacturer uses their own controls that differ in their programming, performance and the achievable path accuracy of the robot. Typical controls are the IRC5 , S4C + ( ABB AG ) and KRC3 ( Kuka AG ).

In addition, there are numerous system houses that bring the industrial robots to life in individual systems that are adapted to the respective customer requirements. In large-scale productions, such as automobile production, often only robots from a single manufacturer are used. This reduces the number of spare parts to be kept in stock . In addition, it is not necessary to train employees on different systems. However, more and more automobile manufacturers are turning to the cheapest robot supplier in order to reduce an overly one-sided robot population and thus the price dependency on a single manufacturer.

Companies like VW , which formerly had their own robot production, have discontinued this with increasing specialization and now obtain their industrial robot requirements externally.

Well-known industrial software products are Kuka Sim (for Kuka), Roboguide (for Fanuc), RoboStudio (for ABB) and Stäubli Robotics (for Stäubli) as well as brand-independent ArtiMinds Robot Programming Suite, Delmia, FASTSUITE, RoboDK, RobotExpert, Robotmaster, SprutCAM Process and Simulate.

Robot density

The robot density, measured in units per 10,000 employees, is a comparative standard for measuring the degree of automation in the manufacturing industry in different countries. The global average global robot density increased from 66 units in 2015 to 74 units in 2016. The average annual growth rate of robot density between 2010 and 2016 was 9 percent in Asia, 7 percent in America and 5 percent in Europe. In a country comparison, the first places in 2016 were occupied as follows: South Korea 631 robots, Singapore 488 robots and Germany 309 robots per 10,000 employees.

Research institutions

Research institutes in German-speaking countries are, in alphabetical order, for example:

Film documentaries

See also


  • Stefan Hesse, Viktorio Malisa (Hrsg.): Pocket book robotics assembly handling . Carl Hanser Verlag, 2010, ISBN 978-3-446-41969-8 .
  • Edwin Kreuzer, Jan-Bernd Lugtenburg, Hans-Georg Meißner, Andreas Truckenbrodt: Industrial robots: technology, calculation and application-oriented design . Springer-Verlag, 1994, ISBN 978-3-540-54630-6 .
  • Alois Knoll, Thomas Christaller: Robotics: Autonomous Agents. Artificial intelligence. Sensors. Embodiment. Machine learning. Service robot. Robots in medicine. Navigation systems. Neural Networks. RoboCup. Architectures . Fischer (Tb.), Frankfurt 2003, ISBN 978-3-596-15552-1 .
  • Wolfgang Weber: Industrial robots. Methods of control and regulation. With 33 exercises . Fachbuchverlag Leipzig, 2002, ISBN 978-3-446-21604-4 .
  • Daniel Ichbiah: Robots. History - technology - development . Knesebeck, 2005, ISBN 978-3-89660-276-3 .

Web links

Commons : Industrial Robots  - collection of images, videos and audio files
Wiktionary: Industrial robots  - explanations of meanings, word origins, synonyms, translations

Individual evidence

  1. 1973 The first KUKA robot  ( page no longer available , search in web archivesInfo: The link was automatically marked as defective. Please check the link according to the instructions and then remove this notice. German, Retrieved May 18, 2008@1@ 2Template: Dead Link /  
  2. ABB technologies that changed the world. Der Industrieroboter, p. 13. (PDF; 3.8 MB) ABB Ltd., accessed on August 31, 2012 .
  3. Functions RT ToolBox2 Software Industrial Robots-MELFA | MITSUBISHI ELECTRIC FA. Retrieved September 27, 2019 .
  4. 2011: The most successful year for industrial robots since 1961 ( memento of the original from February 24, 2014 in the Internet Archive ) Info: The archive link was inserted automatically and has not yet been checked. Please check the original and archive link according to the instructions and then remove this notice. (PDF; 80 kB) (German) accessed on October 21, 2012 @1@ 2Template: Webachiv / IABot /
  5. German companies take advantage of Japan's weakness . Spiegel Online , May 14, 2011
  6. ( Memento of February 27, 2012 in the Internet Archive ) (PDF)
  7. Executive Summary, World Robotics 2019, Industrial Robots. Robot Installations 2018: Now beyond 400,000 units per year. International Federation of Robotics. Retrieved February 12, 2020.
  8. ABB moves robotics HQ to Shanghai . May 27, 2012
  9. Bernd Mewes: IFR: Robot density rises to a new record worldwide. heise online, February 8, 2018, accessed on February 9, 2018 (German).
  10. Press release: Robot density rises to new record worldwide - International Federation of Robotics. February 7, 2018, accessed February 9, 2018 .
  11. ^ IFR: Robot density rises globally. February 7, 2018, Retrieved February 9, 2018 (American English).