Introduction
Robotics is a relatively recent discipline which experienced an incredible growth in the fifty years of its development. Though industrial manufacturing has been and still is the main driver for the development of robotics technologies and theories, service robotics was soon recognized (Engelberger 1989), and it is still today considered, as the main line of development for robotics in the future. Service robotics presently means the application of robotics in services, with the variety of meaning that this can have.
First applications of robots in services have especially been in environments where human access is impossible, dangerous, or difficult. This is the case of space applications, de-mining, underwater tasks. The biomedical field is also an important area of application for robots, in surgery, rehabilitation, and assistance (Menciassi - Laschi 2012).
In summary, however, even with a variety of tasks, contexts, goals, and needs, service robotics aims at providing tools in support of human beings’ activities. Similarly to the wide spreading of ICT technologies in services, robotics as well is going to provide support to human beings extensively.
Robotics and human enhancement
Robotics technologies are then one of the tools for human enhancement. This is intended first of all as the enhancement given by a tool aiding in an activity. It can then assume a stronger meaning when robotics contributes to a real increase of human capabilities. This is the case of robotics aids for recovering human capabilities, in case of aging or disability, as well as the case of wearable robotics for augmenting human strength or endurance.
When these robots are used, issues can emerge on the meaning of being human. We accept that a human being is still a human being when wearing the most common aid we are used to, i.e. a pair of glasses, but what if a person is wearing two leg prostheses, improving his/her running performance, or if he/she is wearing an exoskeleton amplifying his/her strength?
At least, we can see three levels of human enhancement: (i) support in tasks; (ii) aids to recover lost abilities; (iii) augmentation. Such three levels evoke different meanings with respect to the concept of being human, with different acceptability levels:
(i) Robot support in human tasks is usually well-accepted and well-compatible with the common concept of human being. In this context, robots represent tools or appliances, facilitating the accomplishment of tasks by the human user. This is the case, for example, of robots used in professional services, like in space (see the Canadarm on the ISS or the NASA Robonaut performing EVA guided by the astronauts safely inside the ISS), underwater (think of the Titanic exploration missions and the Gulf of Mexico oil spill containment), in surgery (the Da Vinci robot has been used successfully worldwide in approximately 1.5 million various surgical procedures to date) and rehabilitation (the MIT-Manus and the Lokomat robots help in therapy of post-stroke and spinal lesion patients daily), or even in home cleaning (the Roomba vacuum cleaner is the most sold service robot to date). The case of robot vacuum cleaners is probably the easier to convey the idea of enhancement of human activities through supporting/aiding tools. Current lines of research and development in this context are going towards the concept of a ‘Robot Companion’ that can assist its human user in everyday life activities (Dario et al. 2011). This is noticeably in the direction of a personal robot, as a personal assistant.
(ii) In case of disabilities, including those related to aging, robots can become effective aids. Assistive robots for manipulation tasks are already a reality, both as task-specific assistive devices (feeding robots, page turners, and others (Menciassi - Laschi, 2012)), and as general-purpose robots (wheelchair-mounted robot arms (Kim et al. 2012)). In the case of an amputee, upper or lower limb robotic prostheses are currently developed. These assistive robots improve the (motor) capabilities of the users in order to increase their level of activity and ultimately their independence and re-integration. Thus, robotics here contributes to recover lost capabilities, to restore a previously “normal” condition. Of course, contrary to what can be done with biotechnologies, in case of robotics technologies the act of restoration will always be “visible”.
(iii) Augmentation is neither support nor restoration. The starting condition is not a deficiency, but a “normal” condition. The added value of robotics is incremental. Exoskeletons currently in use by the military are intended to make soldiers more powerful.
Such levels may somehow dilute in each other: what if a robotic hand was designed not just to restore but also to augment some of the capabilities of the amputee, for instance by setting higher thresholds to heat and cold? Would that be considered as a form of empowerment? And what if a person would ask to cut one of his/her healthy arm to have the latest version of a robotic hand prosthesis? Would that be considered as a right, as it happens with a transgender? Should the desire to become a cyborg be granted?
However, the question of enhancement, whether support, restoration or augmentation, as well as that of disability, is always relative. In other words, it cannot be absolute because it depends on the context. For instance, a healthy person with an exoskeleton can be an enhanced human in the battlefield but a sort of disabled in the urban environment. Conversely, a person can have a handicap just because of the context.
Impact of robotics on being human
The impact of robotics on the concept itself of being human is definitely among the most disruptive impacts a technology can have on human society. While most questions attain to philosophy, there are also a number of social and legal consequences.
An example can illustrate how and why such issues can emerge in robotics. In 2010, the service robot DustBot, a self-driving vehicles designed to collect garbage bags, drove on the public streets of Peccioli, a small town near Pisa (Italy). It was one of the earliest attempts to integrate a robot in a real human environment by designing the robot for a real service. However, it was necessary to address many societal and legal issues in order to have the robot move on roads (Salvini et al. 2011). From the legal point of view there was the problem of classification according to the Traffic Code. As a matter of fact, the robot could not be defined as a vehicle nor did it fit any of the classification in the Traffic Code. From the social point of view, it was necessary to work with the citizens of Peccioli, by means of public assemblies, in order to avoid forms of resistance, which were given by the fact that the robot was travelling together with cars and pedestrians, and that it was carrying out a public service, instead of human beings.
Designing socially acceptable robots
The social and legal consequences of the application of robots in service raise public issues. However, there are aspects that roboticists themselves take into account when designing robots, in order to preserve core social values and promote an enriching impact on social and societal issues, as opposed to detriment. Not in a priority order, few issues that roboticists consider when designing a robot enhancing human activities are described in the following.
Users’ needs
In service robotics, the design of the robot should take into primary account the needs of the potential users. The robot users are no more trained workers, like in the manufacturing scenario, but users may be common citizens, families, the elderly, or disabled people. It has been demonstrated long ago that the perceived helpfulness is one of the critical drives for acceptability (see the case of a robot for personal assistance to the disabled in Dario et al. 1999).
Robot autonomy
It is possible to distinguish two basic ways in which robotics technologies can enhance human beings. The distinction is based on the robot level of autonomy:
- tele-operated robotics systems (aka telerobotics): robots are operated by a remote user who sends motor commands and receives sensory feedbacks (e.g. images). The interfaces for teleoperation may vary from simple computer input/output devices to master-slave systems such as the Da Vinci surgical robot, or wearable exoskeletons for detecting the operator movements. Within this category there are also hybrid bionic systems, in which the robotic device is connected to the human body and controlled by muscular or nervous signals, like exoskeletons and robotic prostheses, since they are both controlled by the user, even if there is no physical distance between the artifact and the operator.
- Autonomous robotic systems are intended to have some decisional capability, in order to plan and control their own behaviour, based on the current situation, perceived through sensory systems. Common approaches include simple reactive architectures where sensors and actuators are directly connected through simple relations.
- In terms of acceptability, semi-autonomous systems are a good trade-off to find in order to minimize the burden of teleoperating the robot step by step and to maximize the involvement of the user. The user and the robot share the accomplishment of the task to the best benefit of the user.
Teleoperated robotics systems are forms of enhancement usually meant to go beyond the constraints of human body, that is, physical constraints, which can be distance (i.e. drones), scale (i.e. the Da Vinci) or danger (i.e. bomb disposal robots). On the contrary, autonomous systems are forms of enhancement which are meant to overcome the human cognitive as well as bodily limits. The ethical, social and legal consequences of these two basic categories of robotic enhancement are different, since in the former case there is still a connection with a human being while in the latter the robot is separated from the human.
Appearance
Appearance conveys expectations and when designing a robot it represents an important tool for inducing ease of use and acceptability. It has been demonstrated that a human-like appearance, while linearly increasing the acceptability of a machine, can abruptly negatively affect such acceptability (so-called Uncanny Valley, see Mori et al. 2012) when it reaches the level where the machine is perceived as a fake human. On the contrary, the principle of affordance (Norman 1999), i.e. understanding the use of an object from its shape, can be applied in robotics with significant benefit in terms of acceptability.
Soft robotics for robot companions
Soft robotics for a new bodyware
Going back to the concept of service robots for human enhancement, intended as tools that increase human capabilities and performance in given activities, the new challenges of robotics research include a deep re-thinking of the robot bodyware, in addition to robot behaviour. Aiming at robots that can effectively and helpfully accomplish tasks in our daily lives and our environments, we have to consider the profound differences in robot and animal bodies: while robot bodies are based on rigid structures, animal bodies present uncountable elements of compliance, or softness. Bioinspired robotics is taking inspiration from Nature to design robots. In bioinspired robotics the use of soft materials is suggested by the uncountable examples of animal and vegetal systems. Rigid structures, like skeletons, or exoskeletons, are always accompanied by soft tissues. These include mechanisms for varying the material characteristics such as stiffness, elasticity and surface properties, etc. (Kim et al. 2013) for generating motion through muscles and for facilitating sensing in skins through embedded mechanoceptors.
Indeed, the use of soft deformable and variable stiffness technologies in robotics represents an emerging way to build new classes of robotic systems that are expected to interact more safely with the natural, unstructured, environment and with humans, and that better deal with uncertain and dynamic tasks (i.e. grasping and manipulation of unknown objects, locomotion in rough terrains, physical contacts with human bodies, etc.).
Soft robotics for a new artificial intelligence
At the same time, modern views of intelligence, in human and other animals, attribute a stronger role to the physical properties of the body. Beyond the concept of embodiment, i.e. the need of a body for developing intelligence (Brooks 1991), Embodied Intelligence, or Morphological Computation, stresses the interaction with the environment as a key aspect for the emergence of cognition. Embodied Intelligence implies the need for a body with capabilities of interaction with the environment and reaction to environmental changes (Pfeier - Bongard, 2007). The application of Embodied Intelligence in robotics has a great potential for effective advances in real-world applications, in line with the recent development of research in soft robotics, targeting an enriched functionality of soft robots in real-world applications. In fact, according to embodied intelligence, a large part of motor control, which is one of the main objectives of robotics theories and techniques, is accomplished by the bodies itself, by its morphology and by the mechanical properties of its tissues.
Soft robotics progresses and applications
Soft robotics is then now attracting the interest of many robotics researchers worldwide, and it represents one of the current challenges, not only for the development of new components and systems, but also for facing new application perspectives for service robotics. The wide growth of the field of soft robotics is witnessed by many indicators, like the increasing number of scientific publications and conferences, projects and international networks.
An ideal model for both soft robotics and embodied intelligence is the octopus. An investigation on the Octopus vulgaris has led to unveil some of the key principles of embodied intelligence in the octopus and of the octopus dexterity and to their application in the development of soft robotics technologies and of an octopus-like robot (Laschi et al. 2011). According to the biomimetic design principles, a deep study of the animal model has led to identify the key principles of the octopus muscular structure, which has been replicated with a soft material arm structure, actuated longitudinally and transversally by soft actuators (cables and SMA springs). The key principle for octopus variable stiffness has been reproduced, too, with the co-contraction of longitudinal and transverse actuators.
Among the possible applications of the technologies derived from the octopus investigation, are service applications in those fields were robots are requested to do explorations in confined spaces. An example is the use of soft robotics in robotic endoscopy (Cianchetti et al. 2013), where endoscopes can be built that can be soft when entering and navigating inside the human body, but that can increase the stiffness of their parts when needed for local interventions. Another example is the exploration of underwater sites, where soft robots can get in contact with the sea bottom, or reefs, or underwater plants, with no risks of damages (Arienti et al. 2013).
Conclusions
In cases where service robots become tools for explorations of inaccessible environments, as well as in the many other robot services, the relation with the human user, the possible human enhancement, and the consequent impact on being human are very much related to the modality for robot control, specifically in terms of robot autonomy. As described, some level of semi-autonomy represents the best trade-off for maximizing acceptability.
In conclusion, the paradigm that can have a better acceptability is the personal robot assistant, or robot companion: a separate robot, which helps in activities, with a limited degree of autonomy. However, for the personal robotics potential to materialize in people’s life, it is of utmost importance that a suitable legal framework is investigated and put in place.
Acknowledgments
Thanks to Pericle Salvini for discussion and advice. Some of the robotics research works mentioned here have been supported by the European Commission with the projects DustBot, OCTOPUS, and RoboSoft, and by the Fondazione Livorno with the PoseiDRONE project.
Bibliographical references
Arienti A., Calisti M., Giorgio Serchi F. and Laschi C. (2013), PoseiDRONE: design of a soft-bodied ROV with crawling, swimming and manipulation ability, in «Proceedings of the MTS/IEEE OCEANS Conference», San Diego (Calif.), USA, September 21-27, 2013.
Brooks R.A. (1991), New approaches to robotics, in «Science», No. 253, September 1991, pp.1227-1232.
Cianchetti M., Ranzani T., Gerboni G., De Falco I., Laschi C., Menciassi A. (2013), STIFF-FLOP Surgical Manipulator: mechanical design and experimental characterization of the single module, in «Proceedings of the IEEE/RSJ International Conference on Intelligent and Robotic Systems – IROS 2013», Tokyo, Japan, November 4-7, 2013, pp.3567-3581.
Dario P., Verschure P.F.M.J., Prescott T., Cheng G., Sandini G., Cingolani R., Dillmann R., Floreano D., Leroux C., MacNeil S., Roelfsema P., Verykios X., Bicchi A., Melhuish C., Albu-Schäffer A. (2011), Robot companions for citizens, in «Procedia Computer Science», Vol.7, 2011, pp. 47-51.
Dario P., Guglielmelli E., Laschi C., Teti G. (1999), MOVAID: a personal robot in everyday life of disabled and elderly people, in «Technology and Disability», Vol.10, No.2, 1999, pp.77-93.
Engelberger J.F. (1989), Robotics in Service, MIT Press, Cambridge (Mass.) 1989.
Kim S., Laschi C., Trimmer B. (2013), Soft robotics: a bioinspired evolution in robotics, in «Trends in Biotechnology», No.31, pp.287-294.
Kim D.-J., Hazlett-Knudsen R., Culver-Godfrey H., Rucks G., Cunningham T., Portee D., Bricout J., Wang Z., Behal A. (2012), How Autonomy Impacts Performance and Satisfaction: Results From a Study With Spinal Cord Injured Subjects Using an Assistive Robot, in «IEEE Transactions on Systems, Man and Cybernetics, Part A: Systems and Humans», Vol.42, No.1, 2012, pp.2-14.
Laschi C., Cianchetti M., Mazzolai B., Margheri L., Follador M., Dario P. (2012), Soft Robot Arm Inspired by the Octopus, in «Advanced Robotics», Vol. 26, No. 7, 2012, pp.709-727.
Menciassi A., Laschi C. (2012), Biorobotics, in Ziad O. Abu-Faraj (ed.), Handbook of Research on Biomedical Engineering Education and Advanced Bioengineering Learning: Interdisciplinary Concepts, IGI Global, Hershey (Pa.) 2012.
Mori M., MacDorman K.F., Kageki N. (2012), The Uncanny Valley [From the Field], in «IEEE Robotics and Automation Magazine», Vol.19, No. 2, pp. 98-100.
Norman D.A. (1999), Affordance, Conventions and Design, in «Interactions», Vol.6, No.3, pp.38-43, ACM Press.
Pfeifer R., Bongard J.C. (2007), How the Body Shapes the Way We Think: A New View of Intelligence, MIT Press, Cambridge (Mass.) 2007.
Salvini P., Laschi C., Dario P. (2011), Do service robots need a driving license?, in «IEEE Robotics and Automation Magazine», Vol.18, No.2, June 2011, pp.12-13.
E-mail:
