INTRODUCTION
In view of increasing digitalisation, new advances are constantly being made in the field of robotics. Technical system access is increasingly being used in industrial and everyday processes. The aim of the following paper is to compare the capabilities presented in popular science sources as of 2017 with the actual state of the art and research. For this purpose, the field of robots is divided into industrial robots, service robots and social robots.
INDUSTRIAL ROBOTS According to VDI Guideline 2860, industrial robots are universal movement machines with several axes whose movements are freely programmable (i.e., without mechanical intervention) with regard to sequence of movement and
© 2022. Published by AHFE Open Access. All rights reserved. 99
paths or angles and, if necessary, are sensor-guided. They can be equipped with grippers, tools or other manufacturing equipment and can perform handling and/or manufacturing tasks. For Industry 4.0, new approaches are required for safe connections of the robot systems as well as an improvement of their interaction with humans (Kamarul Bahrin MA, et al. 2016).
Deep-Learning According to the secondary source (Cleve T. 2020), it is possible that robots learn together and the learning success is equivalent to the number of robots (what one robot learns in 8 hours - 8 robots learn in one hour). The primary source confirms collective learning at this point, but does not specify how much faster the robots can work as a result.
For example, at CeBIT back in 2017, robot manufacturer Fanuc demon- strated that robots with area vision sensors and deep learning can retrieve parts from a box and not only learn on their own, but also deliver a better result as the number of trials increases. In addition, they store their learning processes in a cloud, which other robots can access simultaneously and both retrieve the information and upload their own learning process.
This also enables robots to work and learn collectively on the same project, where several robots can learn the task correspondingly faster than a single robot (Fanuc 2017). Similarly, the study by Levine et al. 2016 also confirms that robots can show independent learning success in the repeated grasping of different objects. Even though this study only took place under labora- tory conditions and cannot be directly projected to an everyday application, the learning success, especially of hand-eye coordination, cannot be dismis- sed. The same applies to the learning process of grasping different objects at different points. Thus, it could be observed that the system tends to take a different approach when grasping soft objects than when grasping hard obje- cts. For hard objects, the fingers need to be placed on both sides of the object to grasp it successfully. Soft objects, on the other hand, can be grasped by simply pushing into the object, which is most easily achieved by placing one finger in the centre and the other finder on the side of the object (Levine, S. et al. 2016).
Furthermore, the robot manufacturer Boston Dynamics confirms that robots trained by deep learning in this case can precisely locate boxes and pallets even in poor lighting conditions without having to register them beforehand. These robots only need to be trained with a few boxes before- hand and installation and commissioning is correspondingly quick and easy (Boston Dynamics. Pick™2021).
Collaborative, Cooperative or Coexistent?
Especially for industry, collaborative, cooperating and co-existing robots are repeatedly advertised. Prof. Dr.-Ing. Christian Köhler has examined this more closely and has come to the conclusion that robots can only really be descri- bed as collaborative in the rarest of cases. In doing so, he refers to the corresponding taxionomy levels according to Onnasch (Köhler C. 2017). The corresponding excerpt on the taxonomy of cooperation and collaboration
Robots in Popular Sciences Compared with their Real Capabilities 101
from Onnasch: “In contrast to co-existence, cooperation and collaboration describe a genuine collaboration between humans and robots for a common goal achievement. Cooperation involves working towards a higher-level com- mon goal. However, the actions are not directly dependent on each other because there is a clear division of tasks between humans and robots. Humans and robots thus work on different subtasks of the final result, the alloca- tion of which is determined in advance of the task processing. The use of pick-and-place robots in production can serve as an example of this form of cooperation. Both robots and humans work together on a production chain with a common overriding objective: the manufacture of a specific product. For example, the human can label bottles, which the robot then takes over and packs into cartons.
Collaboration describes an interaction as direct cooperation between humans and robots. Both pursue a common goal; in contrast to cooperation, sub-goals are also pursued jointly in this case. This means that partial actions to achieve the goal are also carried out jointly by humans and robots, so that there are immediate coordination challenges. The allocation of subtasks takes place continuously and, if necessary, adapted to the situation directly during the collaboration. Furthermore, collaboration is characterised by the crea- tion and use of synergies” (Onnasch L et al. 2016). This is also shown by an example from a BMW plant published by the robot manufacturer Kuka itself: “Where the workers at the BMW plant in Dingolfing used to have to lift and join heavy bevel gears for front axle transmissions on their own, they now work together with their collaborative colleague, the LBR iiwa [...] This ena- bles close cooperation between humans and robots without any protective fence at all”. Here Kuka describes the work itself as cooperation, but the robot as collaborative. According to Onnasch, this would only be a coopera- tion. However, Kuka also writes that the term “human-robot collaboration” is the most common. However, the “K” in MRK can also stand for both coe- xistence and cooperation, depending on the type of collaboration, and thus refers to Onnasch’s taxonomy (KUKA AG 2021).
In particular, in order for robots to be able to collaborate with humans in a more optimised way, research has already been carried out with the aim of optimising the movement of robots through human intervention. There are already working models for this, but it remains open how a robot is supposed to distinguish which intervention should serve as a learning function for it, or only represents a one-time interruption (Bajcsy A et al. 2017).
In the secondary literature, industrial robots are not found as frequen- tly as other types of robots. This can probably be attributed to the fact that the interested parties tend to be companies that are advertised through other channels. Nevertheless, some publications can be found for the areas of manufacturing as well as maintenance and repair. The largest area of applica- tion today is still manufacturing. However, current robot developments offer promising and potential advantages for the maintenance and repair of indu- strial plants. No deviations from statements from popular science sources can be found.
SERVICE ROBOTS Among a multitude of definitions for the term service robot, the following are frequently cited: “A service robot is a freely programmable motion device that performs services in a partially or fully automated manner. Services in this context are activities that do not serve the direct industrial production of material goods, but rather the performance of services to people and faci- lities (IuG AU 2021). “A service robot is a robot that performs useful tasks for humans or equipment excluding industrial automation application (IFR 2021) The direct benefit of service robots for humans is the main focus. As a subclass of industrial robots, they are again divided into two categories. First, the service robots for domestic, non-commercial use such as vacuum robots or lawn mowers. These are often not monitored and can be controlled by any person. On the other hand, service robots are used for commercial tasks. For example, for commercial applications such as assisted surgery systems or milking robots. These are operated by trained persons. The following is a summary of what state of the art in the field of service robotics is published in popular science sources.
Use of walking robots: There are already different types of service robots, which are used for different service areas. In classical industrial robots, long- term fluid locomotion and manipulation with multiple contact points are problematic to control. Since other robots are not suitable for tasks such as those in the field of aircraft construction, Airbus has joined forces with the French research institute CNRS to advance the technology of walking machi- nes through the laboratory ROB4FAM (Marsiske, H. 2019). As a heavy-duty robot, this walking robot is capable of transporting and attaching large loads. Pulling aircraft weighing 3.3 tons, for example, is also possible at low speed. Climbing and moving over uneven ground is already possible for the heavy- duty robot. Possible errors during movement sequences can be compensated for by speed.
Use in private applications: Consumer robots are finding their place in the private household sector (Vdiv 2021). It is emphasized that there is no such thing as a “jack-of-all-trades robot” that can perform the entire household. There are smaller devices for different tasks that perform or facilitate various steps of everyday life.
Use in nursing and medicine: In the field of nursing, service robots can already completely take over tasks (F.A.Z. 2020). Primarily, they are nee- ded to compensate for the lack of manpower in this service sector. Various telepresence systems, such as the one from the company InTouch Health, perform only a few tasks close to the body, such as listening to the lungs via an attached stethoscope. They allow the physician to make rounds with- out having to be present. The robot, unlike a present, empathic physician, is not able to express emotions, such as compassion, through body language. Robots support human employees in time-intensive tasks by briefly transpor- ting and distributing medications. Medical technology also offers an area of application for service robots (Beck, V. 2021). They provide doctors with more time to concentrate on important tasks by handing out, cleaning and organizing. Interventions assisted by robots are also possible. In this case, a surgeon controls a robotic arm for more precision in execution.
Robots in Popular Sciences Compared with their Real Capabilities 103
Use in gastronomy: In the restaurant industry, robots are used for serving or for cooking or preparing simpler operations. These speeds up the process and allows a restaurant to expand its capacity.
Use in tourism: The use of service robots is expected in the tourism industry (DER SPIEGEL 2016). The robots “Pepper”and “Mario”are in use on cruise ships of the Aida Cruises fleet. They can recite daily offers and excursion tips or act as aids at the reception desk. These robots cannot handle abrupt changes of topic, unrelated questions and irony. Therefore, they offer no sub- stitute for a human receptionist. However, they are considered an interesting experience for some visitors.
SOCIAL ROBOTS Social robots are primarily used in the field of health care or public admini- stration. While they are used in the health sector for nursing activities and therapeutic measures, the focus of robots in public administration is on com- munication with humans. In the course of the project study, the propagated application possibilities and prognosis of the use of different types of robots are looked at. These are then compared with the actual evaluations from the corresponding primary sources, expert interviews and statistics. For this purpose, the following picture emerged in social robotics. Out of 34 articles on social robotics, 15 articles dealt with the possible uses of social robotics in healthcare. This suggests that the most important field of application for social robotics in the future will be in the area of care and health. Many reports propagate a coming nursing crisis. The number of people in need of care will increase steadily, while the already thin staffing situation will be confronted with problems as a result. According to an article in the FAZ, in 2035 there will be a shortage of about 500,000 skilled workers in the health sector who can look after and care for people in need of care (F.A.Z.a2021). If one looks at studies on the future occupancy of nursing homes, the sta- tements are fundamentally true (Radtke 2020). Regardless of the skill and effectiveness of social robots, the technology can only enter the market if it is accepted by the carers and the people being cared for. Experts in the field of nursing generally have a higher acceptance level for the use of technologies and robots in the field of nursing. However, the acceptance level for activi- ties with a higher interpersonal component is lower. Particularly striking in the following table is the low level of acceptance for everyday conversations. Social and emotional activities, and thus the integration of social robots into everyday care, are therefore rated particularly critically. The predominant opinion of secondary science articles towards social robotics considers the use of these technologies ethically and morally questionable to indefensible (F.A.Z.b2021) (DW 2019).
The German Ethics Council has published a statement specifically on this discussion and issued a recommendation for the implementation of robotics in care. It does not come to the conclusion that the use of robots in care is ethically and morally questionable. Robotics in care can help in the future, as long as it is “aligned with the goals of good care and assistance” (Deutscher Ethikrat 2020). It should therefore be tested in a practical way already in
the development phase. The ethical evaluation of robotics thus depends on the way it is implemented and cannot be presented as unjustifiable from the outset.
The predominant opinion of secondary science articles towards social robotics considers the use of these technologies ethically and morally que- stionable to indefensible (F.A.Z.b2021) (DW 2019).
The German Ethics Council has published a statement specifically on this discussion and issued a recommendation for the implementation of robotics in care. It does not come to the conclusion that the use of robots in care is ethically and morally questionable. Robotics in care can help in the future, as long as it is “aligned with the goals of good care and assistance” (Deutscher Ethikrat 2020). It should therefore be tested in a practical way already in the development phase. The ethical evaluation of robotics thus depends on the way it is implemented and cannot be presented as unjustifiable from the outset.
CONCLUSION
Only minor discrepancies between the secondary literature studied and the state of the art can be observed in the texts examined.
For the industrial robots, no deviations from statements from popular sci- ence sources can be found. In the secondary literature, however, these are not found as frequently as the other robot types studied. This is probably since the interested parties tend to be companies that are advertised through other channels. Nevertheless, some publications can be found for the manufactu- ring and maintenance and repair sectors. The largest application area today is still manufacturing. However, current robot developments offer promising and potential benefits for the maintenance and repair of industrial plants.
In comparison, the research on service robots have shown that a variety of characteristics and capabilities are attributed to them in the secondary literature, most of which are consistent with the status quo of service robotics.
Service robots are used in everyday processes alongside production and manufacturing processes. However, they are not expected to be able to fully take over human activities, especially those that require emotional responses.
It can be noted that the state of the art in service robots has advanced worl- dwide. They are increasingly being used in areas such as catering and nursing. The state of the art is that further research is needed on items such as estima- tion or flexibility to provide mature results. Robotic systems are increasingly able to take on non-standard tasks that were previously left to humans, at an economically viable cost. The increasing intelligence of machines, as well as their memory and learning capabilities, will enable interaction not only between humans and machines, but also between machines and systems. Col- laborative work is increasingly enabled by optical, acoustic and haptic signal processing systems. This allows robots to work better with humans without endangering them in their workspace. The safety of humans inside is the most important prerequisite for robots in the service sector, as they have to manage without a protective cage, especially in areas without a standardized working environment (Rethink Robotics 2021).
Robots in Popular Sciences Compared with their Real Capabilities 105
In the next area examined, social robotics, the claims from the secondary literature, as with the previous robot types, deviate little from the primary scientific facts. In the technical area, the claims of the secondary literature are fundamentally true, although the ability of robots is generalized in some aspects. For example, robots are said to be able to recognize human emoti- ons, even if this is limited to only a few emotions. Especially in the ethical area, the statements of the secondary literature differ from the primary scien- tific works. Many popular scientific articles question the use of social robots for nursing activities and consider them morally questionable. However, the report of the German Ethics Council sees the inclusion of social robots as possible and even helpful, provided they are used sensibly and thoughtfully.
REFERENCES
Bajcsy A et al. (2017) Learning Robot Objectivesfrom Physical Human Interaction.
Handbook of human systems integration. New Jersey: Wiley. Beck, V. (2021): Roboter im OP: Wo sie eingesetzt werden und was sie können, https://www.produktion.de/technik/roboter-im-op-wo-sie-eingesetzt-we rden-und-was-sie-koennen-337.html Boston Dynamics. Pick™(2021) https://www.bostondynamics.com/pick Cleve T. (2020) Moderne Robotik: Eine kurze Entwicklungsgeschichte – und ein Ausblick. https://cloudblogs.microsoft.com/industry-blog/de-de/manufacturi ng/2020/01/30/moderne-robotik-entwicklungsgeschichte-ausblick/. DER SPIEGEL (2016), Hamburg, Germany, 2016, https://www.spiegel.de/reise/akt
uell/tourismus-trend-gestatten-ich-bin-ihr-service-roboter-a-1082235.html Deutscher Ethikrat (2020). Robotik für gute Pflege” – Hintergrund und Positionen des Deutschen Ethikrats. JuKiP - Ihr Fachmagazin für Gesundheits- und Kinderkrankenpflege, 09(05), 49–52. https://doi.org/10.1055/a-1218-7999. DW (2019). Was sollen Roboter in der Pflege leisten? https://www.dw.com/de/was-
k%C3%B6nnen-was-sollen-roboter-in-der-pflege-leisten/a-49365268. F.A.Z. (2020): Navigationshilfe für Empathie, https://www.faz.net/aktuell/wissen/m edizin-ernaehrung/die-pflegeroboter-kommen-eine-navigationshilfe-fuer-empath ie-16679061. F.A.Z.a(2021) Die Robotik entdeckt das Alter. FAZ.NET. https://www.faz.net/aktu
ell/karriere-hochschule/pflegeroboter-die-robotik-entdeckt-das-alter-17176850. html. F.A.Z.b(2021). Die Robotik entdeckt das Alter. https://www.faz.net/aktuell/karriere -hochschule/pflegeroboter-die-robotik-entdeckt-das-alter-17176850.html. Fanuc (2017). Roboter arbeiten bei Deep Learning zusammen; https://www.fanuc.
eu/de/de/wer-wir-sind/news-and-events/de-cebit-2017 International Federation of Robotics (2021), in: IFR International Federation of Robotics, o. D., https://ifr.org/service-robots IuG AU (2021) http://www.informatik.uni-oldenburg.de/%7Eiug15/au/service_rob
oter.html#SK_ Köhler C. (2017) Zusammenarbeit von Mensch und Roboter. https: //www.ingenieur.de/fachmedien/vdi-z/automatisierung-vdi-z/zusammenarbeit -von-mensch-und-roboter/. Kamarul Bahrin MA, et al. (2016). Industry 4.0: A Review on Industrial Automation and Robotic. Jurnal Teknologi 2016; 78(6-13).
KUKA AG (2021) Mensch-Roboter-Kollaboration bei der Arbeit - KUKA AG; 2021.000Z. URL: https://www.kuka.com/de-de/future-production/mensch-robo ter-kollaboration. Levine, S. et al. (2016) Quillen D. Learning Hand-Eye Coordination for Robotic
Grasping with Deep Learning and Large-Scale Data Collection. Marsiske, H. (2019): Wann kommen Laufroboter in der richtigen Welt?, in: heise online, https://www.heise.de/newsticker/meldung/Wann-kommen-Laufroboter-in -der-richtigen-Welt-4431016.html Onnasch L et al. (2016) Mensch-Roboter-Interaktion - Eine Taxonomie für alle
Anwendungsfälle. Bundesanstalt für Arbeitsschutz und Arbeitsmedizin. Radtke (2020). Prognostizierter Bedarf an stationären und ambulanten Pflegekräften in Deutschland bis zum Jahr 2035. Statista. https://de.statista.com/statistik/daten/ studie/172651/umfrage/bedarf-an-pflegekraeften-2025/. Rethink Robotics (2021) Sawyer Collaborative Robots for Industrial Automation
https://www.rethinkrobotics.com/de/sawyer Vdiv (2021): Serviceroboter übernehmen immer mehr Aufgaben, https://www.ingeni eur.de/technik/fachbereiche/robotik/serviceroboter-erobern-heimischen-herd/