These innovations make industrial robots significantly better

Robots can now be used for applications that previously could only be done manually. Not only AI and sensors play the main role.

Automated processes are now possible in robotics, above all thanks to AI and image processing systems, which previously could only be carried out manually
industrial robots

The possibilities of automation are constantly expanding and are moving into focus for robot applications that could previously only be carried out manually. Sorting waste on a conveyor belt, for example. And just as in the past two years with solutions for automated welding has developed into a large market, automated screwdrivers will now also automate many processes in assembly and, above all, make them more precise.
Among other things, in addition to being equipped with AI, image processing systems and sensors, the new robot generations are easier to install, program and network – the best prerequisites for combining traditional production with digital strategies. Below you can find out which six components of the robot improve your application or make it possible in the first place.

Despite their compact design, the 3FG15 grippers have a load capacity of 15 kg
industrial robots


Grippers are becoming increasingly intelligent in order to be able to communicate with their technical environment via IO-Link, Profinet, EtherCat or EtherNet/IP for sensitive use. Not only do they receive information, the sensor data from a gripper, such as a temperature or acceleration, can also be sent to the system controller or stored in the data cloud.
The vacuum gripping systems from Schmalz, for example, provide important data on the condition of the system and enable functions such as condition monitoring and predictive maintenance. A comprehensive IO-Link configuration file IODD (IO Device Description) is available to the user for this purpose, which includes information on identification, device parameters, process and diagnostic data as well as communication profiles. The exchange of devices is also simplified since the IO-Link protocol contains an automatic data transfer mechanism.

Integration and operation significantly simplified

To simplify integration and operation, grippers are increasingly being designed as plug-and-produce solutions that are fully operational right out of the box on robots. This is made possible, for example, by the standard integration of a quick-change system in its tools (Quick Changer) and by providing a uniform programming logic (One System – Zero Complexity). Schmalz is also increasingly supplying software plugins that guide the user through the configuration with graphical support.
The greater compatibility of the grippers with robot arms from different manufacturers also makes integration easier for users and also expands the possible applications. New areas of application always result in new approaches to handling components. With the help of adhesive forces, such as with the new Adheso gripper technology from Schunk, sensitive components can be gripped completely energy-free, gently and without leaving any residue.


In order to extend the possible uses of an industrial robot to work environments in which potential contamination, for example from hydraulic oils, is excluded, pneumatic/compressed air-based grippers are replaced by electrically operated actuators in the pharmaceutical or food industry. Grippers with an electric motor reduce maintenance costs and simplify handling since they do not require external hoses that are prone to wear.
Drives that are compatible with intelligent interfaces such as RS232 or CAN enable communication between a gripper and the higher-level controller for the purpose of grip monitoring and gripping part recognition as well as flexible adjustment of gripping force and gripper stroke. Maintenance-free drives with high reliability and performance at the lowest possible weight are particularly important for lightweight robots/robots asked. The reason: the lower the weight of the EOT (End of Arm Tool), the more load a robot can lift.

Compact designs make jobs easier for grippers

More compact designs also reduce the interfering contours of the gripper. Faulhaber’s BXT series external rotor motors are an example of a unique ratio of torque to weight and construction volume. The iron-core motors with 14 high-performance rare-earth magnets on the rotor and 12 teeth on the stator are only 14 mm, 16 mm and 21 mm long, respectively, and are therefore suitable for applications that require a short drive solution with a high torque of up to 134 mm require.


Cobots are becoming more and more sensitive and are now also taking on tasks in precision assembly or surface processing, such as deburring, grinding or polishing; the sensors can be integrated into the gripper itself or mounted as an external module on the tool flange of the robot arm.
Sensor technology right at your fingertips, for example, gives OnRobot’s RG2-FT grippers so much intelligence that they recognize the exact workpiece position. The previous programming of the parameters is therefore not necessary. The e-Series models from Universal Robots, on the other hand, have a force-torque sensor integrated into the tool flange: The highly sensitive sensor registers even the smallest of movements and gives the cobot a delicate touch.

Vision Systems

Artificial intelligence helps to optimize performance through algorithms and self-learning systems, from which vision solutions such as camera-supported bin picking benefit in particular. The ready-to-connect bin picker SBPG from Schmalz has proven itself for automated bin picking. It works in combination with a 3D camera system that performs visual part recognition allowed. Weighing only a few hundred grams, the bin picker reliably picks workpieces out of the container – regardless of whether they are stored randomly or pre-sorted.
More and more suppliers are bringing image processing systems onto the market, with the help of which robots capture their surroundings visually and multi-dimensionally. This allows them to recognize objects even if they are stacked or overlapping. Vision systems such as On Robots Eyes allow robots to visually recognize parts and perceive their surroundings spatially. This allows them to pick and pick in a more targeted manner.

AI and deep learning in action

Here, too, AI and deep learning are used in some cases: UR+ partners from Universal Robots have developed a sensor-based vision solution with which UR cobots master ‘bin picking’. Users no longer learn the robot for individual objects, but for the gripping process itself.
Equipped with a camera, gripper and appropriate software, a cobot can recognize any object regardless of coordinates and independently identify optimal gripping points. An AI vision sensor integrated into the camera is constantly learning, so that the process runs more and more smoothly – it is not necessary to teach in the components to be gripped.

Robot controllers from the Omni-Core family offer the necessary flexibility to integrate the latest digital technologies
Robot controllers


New control strategies based on AI – more specifically on machine learning – make it possible to automate manual production steps. For example, the AI-powered robotic controller Mirai from Micropsi Industries extends the robot’s native control and gives it the valuable ability of eye-hand coordination.
In contrast to classic vision systems, in which a camera remeasures every situation and every workpiece, AI-driven controls react to variances in real-time in the environment and continuously steer the robot through its path. This is made possible by the interaction of camera and software: In the future, robots can increasingly be ‘trained’ with AI-based solutions instead of being programmed.

Positioning tasks faster and easier to solve

Mirai has also recently acquired what is known as positioning skills, with which positioning tasks can be solved even more quickly and easily. To do this, the human employee guides the robot to the target and then ‘shows’ it the area around the target with the camera, whereby the robot ignores its own path. The AI ​​​​control then independently decides on the best path and guides the robot through its path each time anew.
The operation of collaborating robots is also becoming increasingly easier, such as with the handy Teach Pendant control unit from UR. The intuitive user interface also enables users without previous knowledge to control the robot and adapt the respective application to the situation. Operator panels of the latest controller generations also offer enormous flexibility thanks to the hot-plug function, for example. In this way, the tablet can be disconnected during operation and connected to another controller.

Connectivity to the digital platform and cloud solutions

Customization options are also flexible, as are the diverse assembly options that result from smaller footprints. One of the most important control features is the connectivity to digital platforms and cloud solutions, which can be used to improve the performance and reliability of individual robots through to entire robot fleets.
The digital ABB Ability platform includes networked services as well as the ABB safety solution SafeMove. With that, the controls offer basic functions that can enable collaboration with a connected industrial robot.

Human-Robot Collaboration

‘Hybrid’ collaborative robots are full-fledged industrial robots that can work at high speed but fall back to a safely reduced speed as soon as the human is directly in the workspace. After all, permanent human-robot interaction is only required for very few automation tasks.
In fact, in many cases, the human is rather ‘in the way’ of the robot since the robot has to work slowly – at a safely limited speed – in direct contact with its operator, which leads to long cycle times. Yaskawa’s Motoman HC models, for example, are such hybrid robots with a handling weight of 10 or 20 kg. The required safety in direct contact with the operator is guaranteed by its power and force limit circuit board, which enables flexible interaction between the robot and its environment.

Classic robot systems also work without a protective fence

However, classic robot systems can also work without a protective fence, provided they are equipped with a safe robot controller. The only difference to an HRC-capable hybrid robot is that the classic industrial robot has to stand still in the presence of humans, while the collaborative hybrid robot continues to work slowly and safely. Standard safety technology (such as safety laser scanners, safety curtains or safety mats) can be used to detect whether a person is present or not.


Automation beginners do not ask for individual components, but for applications. Here application kits, Which means a standardized component bundle with supplementary software services, solutions that are optimally coordinated with each other and with the robot. In addition, online platforms network end-users, system integrators and component manufacturers and offer standardized, attractively priced robot cells that automate processes that were previously carried out manually with little installation effort. With the CoLab, Schunk offers an essential building block for these platforms, because this is where these applications are validated and provided with the right components and software building blocks.