Ethically Motivated Recommendations for EU Policies and Regulations concerning Robotics and its Convergence with Bionic and Software Agent Technologies.

Full text

Deliverable D6
SAS 6 - 017759
ETHICBOTS
Emerging Technoethics of Human Interaction with Communication,
Bionic and Robotic Systems
Coordination Action
Structuring the European Research Area
D6: Ethically Motivated Recommendations for EU Policies,
Regulations, and Further Research concerning Robotics and its
Convergence with Bionic and Software Agent Technologies
Due date of deliverable: April 30th, 2008
Actual submission date: May 5nd, 2008
Start date of project: 1 November 2005 Duration: 24 months + 6 months extension
University “Federico II”, Naples
Revision: Draft
Project co-funded by the European Commission
Dissemination Level
PU
Public
PP
Restricted to other programme participants (including the Commission Services)
x

RE
Restricted to a group specified by the consortium (including the Commission Services)
CO
Confidential, only for members of the consortium (including the Commission Services)

Project funded by the European Community as Coordination Action
Contract SAS 6 Nr. 017759
Emerging Technoethics of Human Interaction
with Communication, Bionic and Robotic
Systems
Deliverable D6: Ethically Motivated Recommendations for EU
Policies, Regulations, and Further Research concerning Robotics
and its Convergence with Bionic and Software Agent Technologies
Editors:
Rafael Capurro, Thomas Christaller, Guglielmo Tamburrini
Co-author(s):
Rafael Capurro, Thomas Christaller, Edoardo Datteri,
Alessandro Giordani, Cecilia Laschi, Michael Mock,
Michael Nagenborg, Fiorella Operto, Pericle Salvini,
Guglielmo Tamburrini, Gianmarco Veruggio, Jutta Weber

Deliverable D6
Page 4/29
Status: Final
Version: 7
Date: 30.04.08
Revision history
Deliverable administration and summary
Project acronym: Ethicbots
ID: SAS-6-017759
Document identifier:
Ethicbots-D6
Leading partner: University “Federico II” of Naples
Report version: 8
Report preparation date: XXXXXX
Classification: Confidential
Nature: Deliverable
Author(s) and contributors:
Status:
Plan
Draft
Working
X
Final
Submitted
Approved
The Ethicbots Consortium has addressed all comments received, making changes as
necessary. Changes to this document are detailed in the change log table below.
Date
Edited by
Status
Changes made
March
4th, 2008
Jutta Weber, Gianmarco
Veruggio
Draft
Care robots, military uses of robots
March, 10,
2008
Michael Nagenborg
Edutainment robots, robots in the workplace,
relevant legislation
April 7th,
2008
Thomas Christaller,
Michael Mock, Fiorella
Operto, Gianmarco
Draft
Contributions on learning robots, revisions on
edutainment robots, usage of intentional language

Deliverable D6
Page 5/29
Status: Final
Version: 7
Date: 30.04.08
Veruggio
April 21st,
2008
Guglielmo Tamburrini
Draft
First complete draft
April 28th,
2008
Rafael Capurro
Draft
Revised complete draft
April 28th
Edoardo Datteri
Draft
Addition: Surgery robotics
April 30th
Cecilia Laschi, Pericle
Salvini
Draft
Additions on care robots, surgery, and brain-robot
communication
April 30th
Editors
Draft
Final revisions
Notice that other documents may supersede this document. A list of latest public Ethicbots
deliverables can be found at the Ethicbots information server.
Copyright
This report is © Ethicbots Consortium 2008. Its duplication is restricted to the personal use
within the consortium, funding agency and project reviewers.
Citation
Rafael Capurro, Thomas Christaller, Guglielmo Tamburrini (editors), Rafael Capurro,
Thomas Christaller, Edoardo Datteri, Alessandro Giordani, Cecilia Laschi, Michael Mock,
Michael Nagenborg, Fiorella Operto, Pericle Salvini, Guglielmo Tamburrini, Gianmarco
Veruggio, Jutta Weber (authors), Deliverable D6 – Ethically Motivated Recommendations for
EU Policies, Regulations, and further Research concerning Robotics and its Convergence
with Bionic and Software Agent Technologies. Ethicbots Consortium, c/o University “Federico
II” of Naples
Acknowledgements
The work presented in this document has been conducted in the context of the EU
Framework Programme No. 6 and is funded by the European Commission. Their support is
appreciated.

Deliverable D6
Page 6/29
Status: Final
Version: 7
Date: 30.04.08
The partners in the project are:
University "Federico II", Naples, Physical Science Department and Department of Computer
and Systems Engineering, Italy (coordinator)
Fraunhofer Institute for Autonomous intelligent Systems, Sankt Augustin, Germany
Scuola di Robotica, Genova, Italy
Institute of Applied Philosophy, Faculty of Theology, Lugano, Switzerland
University of Reading, department of Cybernetics, UK
Hochschule der Medien University of Applied Sciences, Stuttgart, Germany
LAAS-CNRS, Toulouse, France
Scuola Superiore Sant'Anna, Pisa, Italy
University of Pisa, Department of Philosophy, Italy
Middlesex University, Interaction Design Centre, School of Computing, London, UK
More information
Public Ethicbots reports and other information pertaining to the project are available through
Ethicbots public information service under http://ethicbots.na.infn.it.

Deliverable D6
Page 7/29
Status: Final
Version: 7
Date: 30.04.08
Table of contents
EXECUTIVE SUMMARY ................................................................................................................................... 7
PREAMBLE ........................................................................................................................................................... 8
GOALS OF THE ETHICBOTS PROJECT ................................................................................................................. 8
SCIENTIFIC AND TECHNICAL BACKGROUND ........................................................................................................ 9
TECHNOLOGY FORESIGHT BACKGROUND ............................................................................................................ 9
FRAMEWORK PROVIDED BY INTERNATIONAL PRINCIPLES AND CHARTERS ......................................................... 10
SCOPE OF THE PRESENT RECOMMENDATIONS ..................................................................................................... 12
ROBOTIC TECHNOLOGIES: ETHICAL MONITORING AND RECOMMENDATIONS .................... 14
LEARNING ROBOTS ............................................................................................................................................. 14
STABILITY AND UNCERTAINTY .......................................................................................................................... 14
POSITION OF HUMANS IN THE CONTROL HIERARCHY .......................................................................................... 14
ROBOTIC ACTION AND ETHICALLY MOTIVATED DESIGN STANDARDS ................................................................. 15
PRIVACY ............................................................................................................................................................ 15
RACE, CLASS, GENDER, INTENTIONAL LANGUAGE ............................................................................................. 16
RECOMMENDATIONS ABOUT DOMAIN-SPECIFIC HUMAN-ROBOT INTERACTIONS ................ 17
CARE ROBOTS .................................................................................................................................................... 17
EDUTAINMENT ROBOTS ...................................................................................................................................... 18
ROBOTIC SURGERY ............................................................................................................................................. 19
REPLACING HUMANS IN THE WORKPLACE .......................................................................................................... 20
REMOTE CONTROL AND COOPERATION IN THE WORKPLACE ............................................................................... 20
MILITARY USES OF ROBOTS ................................................................................................................................ 21
CONNECTING THE HUMAN BRAIN TO ROBOTS .................................................................................... 23
IMPLANT TECHNOLOGIES FOR CONNECTING NERVOUS SYSTEMS TO ROBOTS ...................................................... 23
NON-INVASIVE BCIS AND ROBOTIC SYSTEM CONTROL ...................................................................................... 23
AUTONOMY ........................................................................................................................................................ 24
RESPONSIBILITY ................................................................................................................................................. 24
DISTRIBUTIVE JUSTICE ....................................................................................................................................... 25
PRIVACY ............................................................................................................................................................ 25
PERSONHOOD ..................................................................................................................................................... 25

Deliverable D6
Page 8/29
Status: Final
Version: 7
Date: 30.04.08

Deliverable D6
Page 9/29
Status: Final
Version: 7
Date: 30.04.08
Executive summary
Ethical issues arising from the interaction of humans with robots have
been the main focus of the ETHICBOTS FP6 Science and Society
Coordination Action. In view of the convergence and increasing
integration of robotics with bionics and software agent research, the
ETHICBOTS core business reaches out to include selected ethical
issues arising from the interaction of humans with softbots and bionic
devices. Indeed, IC technologies enable robots to interact remotely with
software agents hosted by other robots or computer systems; and robotic
systems are embedded as mechatronic components of bionic systems
which are formed by both biological and machine parts.
In the ETHICBOTS project, the promotion and the protection of
fundamental rights have provided the chief coordinates for the analysis
of ethical issues and the shaping of related recommendations.
This report presents ethically motivated recommendations for further
research about the ETHICBOTS domain of inquiry and for the
development of policies and regulations concerning robotic technologies
and their integration with technologies in bionics and softbot AI research.
Recommendations about robotic technologies in general concern
learning robots, stability and uncertainty issues in robotics, position of
humans in the control hierarchy, robotic action legibility, privacy, race
and class issues, gender issues, and misleading uses of intentional
language to describe robot behaviour.
Recommendations about domain-specific human-robot interactions
concern care robots, edutainment robots, robotic surgery, robot in the
workplace, and military uses of robots.
A final section is devoted to ethical issues concerning bionic
technologies for interfacing human brains to robots.

Deliverable D6
Page 10/29
Status: Final
Version: 7
Date: 30.04.08
Preamble
Goals of the ETHICBOTS project
Ethical issues arising from the interaction of humans with robots have
been the main focus of the ETHICBOTS FP6 Science and Society
Coordination Action (1 November 2006 through 30 April 2008).
In view of the convergence and increasing integration of robotics with
bionics and software agent research, the ETHICBOTS core business
reaches out to include selected ethical issues arising from the interaction
of humans with softbots and bionic devices. Indeed, IC technologies
enable robots to interact remotely with software agents hosted by other
robots or computer systems; and robotic systems are embedded as
mechatronic components of bionic systems which are formed by both
biological and machine parts.
Interactions of human beings with robotic systems - either in isolation or
in combination with softbot and bionic devices - notably include shared or
delegated action control. On the part of the robotic system, action control
involves sensori-motor coordination capabilities, in addition to varying
combinations of machine perception, learning, reasoning, planning, and
communication capabilities.
Distinctive ethical issues arise from sharing action control with or
delegating it to machines possessing some of the above sensori-motor
and cognitive capabilities. The ETHICBOTS consortium has pursued the
prior goal of identifying and analysing novel ethical issues arising in this
special human-machine interaction context.
Crucial enabling factor for this analytical undertaking has been the
pluridisciplinary character of the group of researchers forming the
ETHICBOTS consortium, whose range of expertise spans robotics,
artificial intelligence, anthropology, moral philosophy, philosophy of
science, cognitive science, gender studies, and science & technology
studies.
In the ETHICBOTS project, the promotion and the protection of
fundamental rights are the chief dimensions along which ethical issues

Deliverable D6
Page 11/29
Status: Final
Version: 7
Date: 30.04.08
have been analysed. The focus on the promotion and protection of
fundamental rights has enabled the consortium to achieve a
comprehensive ethical perspective which carries with it significant policy
making implications.
This report presents ethically motivated recommendations for further
research about the ETHICBOTS domain of investigation and for the
development of policies and regulations concerning robotic technologies
and their integration with technologies in bionics and softbot AI research.
These recommendations are grounded into the analysis of both ethical
issues and their broad policy making implications carried out by the
ETHICBOTS working group.
Scientific and Technical Background
For the scientific and technical underpinning of the recommendations
presented in the ensuing sections, the reader is referred to the other
deliverables released by the ETHICBOTS consortium, and in particular
to
D1: Analysis of the State of the Art in Emerging Technologies for the
Integration of Human and Artificial Entities
D2: Methodology for the Identification and Analysis of Techno-Ethical
Issues
D4: Analysis of National and International EU regulations and Ethical
Councils Opinions Related to Technologies for the Integration of Human
and Artificial Entities
D5: Techno-Ethical Case-Studies in Robotics, Bionics, and Related AI
Agent Technologies
Technology foresight background
A triage of technologies, projects and systems was carried out by the
ETHICBOTS consortium in order to distinguish issues of rights protection
and promotion that concern society now from issues that may possibly
concern society in a fairly distant future only. This triage was carried out

Deliverable D6
Page 12/29
Status: Final
Version: 7
Date: 30.04.08
applying imminence, novelty, and potential social pervasiveness
selection criteria (see Ethicbots deliverable D2, p. 42ff). It is worth noting
that the ETHICBOTS consortium, by endorsing these selection criteria,
has opted for attending in a marginal way only to issues concerning
attribution of rights and responsibilities to robots and to issues reflecting
broad transhumanist views and aspirations.
Another temporal refinement discussed and introduced by the
ETHICBOTS consortium concerns the distinction between long-term
visions driving research in robotics and the near-term perspectives of
robotics research and technology transfer (see deliverable D5). Long-
term visions are taken into account in this report chiefly by means of
ethical monitoring recommendations, which mostly point to the need for
further ethical inquiry and bear no direct regulatory implications.
The need to distinguish between long-term vision and near-term
perspectives for the purpose of issuing recommendations about policies
and regulations is particularly evident in the case of service and personal
robotics, which presently attract considerable attention from general
public and the media. A long-term goal of service and personal robotics
is to enable rich and flexible human-robot interactions in homes, offices,
and other environments that are specifically designed for human
activities. Results obtained in this rapidly growing area of research are
impressive when gauged by the yardstick of scientific and technological
advancement. Their near-future practical significance, however, is more
difficult to assess. Indeed, near future projections licensed by robotic
demonstrations concern restricted forms of cooperative behaviour. And
major theoretical and technological problems have to be solved before
deft interactive robots will step out of research labs and will be ushered
in our homes. (See Deliverables D1, D2, and D5.)
Framework provided by international principles and charters
Crucial for the interpretation of fundamental rights promotion and
protection in the context of robotics and related softbot and bionic
technologies are the articles included in the six sections of the Charter of
Fundamental Rights of the European Union of 28 September 2000,
which concern the promotion and protection of dignity, freedoms,
equality, solidarity, citizenship, and justice.

Deliverable D6
Page 13/29
Status: Final
Version: 7
Date: 30.04.08
In addition to this, significant to ETHICBOTS consortium analyses and
recommendations have been the following:
The European Union Treaty and in particular Article 6 of the common
provisions concerning the respect for fundamental rights;
Directive 2002/58/EC of the European Parliament and of the Council of
12 July 2002 concerning the processing of personal data and the
protection of privacy in the electronic communication sector;
The Principles of Good Governance stated in the European Governance
White Paper issued by the EC on 25 July 2001, and in particular the
principles of openness, participation, effectiveness, and coherence;
The Opinion “Ethical aspects of ICT implants in the human body”
adopted by the European Group on Ethics in Science and Technology
(EGE) on 16 March 2005;
The Declaration of Principles of the World Summit on the Information
Society of 12 December 2003, in particular of Article 58 on the use of
ICTs and Article 59 on the abusive uses of ICTs;
The Council of Europe Convention for the Protection of Human Rights
and Dignity of the Human Being with regard to the Application of Biology
and Medicine: Convention on Human Rights and Biomedicine, signed on
4 April 1997 in Oviedo;
The United Nations Convention on the Rights of the Child. Adopted and
opened for signature, ratification and accession by General Assembly
resolution 44/25 of 20 November 1989 entry into force 2 September
1990, in accordance with article 49;
The United Nations Convention on the Rights of Persons with
Disabilities. Adopted on 13 December 2006 during the sixty-first session
of the General Assembly by resolution A/RES/61/106;
The Geneva Conventions and Additional Protocols.
In the context of specific areas of applications additional documents have
been considered, which included the Council Directive of 12 June 1989

Deliverable D6
Page 14/29
Status: Final
Version: 7
Date: 30.04.08
on the introduction of measures to encourage improvements in the safety
and health of workers at work (89/391/EEC), the United Nations United
Nations Convention on prohibitions or restrictions on the use of certain
conventional weapons which may be deemed to be excessively injurious
or to have indiscriminate effects (1980), the United Nations Optional
Protocol to the Convention on the Rights of the Child on the sale of
children, child prostitution and child pornography (2000), and the
Recommendation of the European Parliament and of the Council of 20
December 2006 on the protection of minors and human dignity and on
the right of reply in relation to the competitiveness of the European
audiovisual and on-line information services industry (2006/952/EC).
An overview of the ethical regulations taken into account within the
ETHICBOTS project was presented in Annex to Deliverable D4.
Scope of the present recommendations
Consistently with the goals of the ETHICBOTS project, these
recommendations focus on shared or delegated action control arising in
interactions of human beings with robotic systems. Softbot and bionic
devices are taken into account here insofar as they may be combined
with robotic devices. Accordingly, this report is not concerned with
softbot technologies in general, as these are developed in Artificial
Intelligence research, or with bionic technologies in general, as these are
developed in bioengineering research.
These recommendations address ethical issues, but are in no way
confined to possible or actual threats to fundamental rights posed by
robotic and related softbot and bionic technologies. Indeed, a central
goal of ETHICBOTS has been that of encouraging, on the basis of an
ethically motivated reflection, responsible scientific and technological
research. Accordingly, the potential of robotic and related technologies to
promote fundamental rights is emphasized throughout the present report.
The ETHICBOTS consortium did not address explicitly the intercultural
dimensions of robotics and related bionics and AI technologies.
However, the consortium participants deem that intercultural issues
play, in a globalized world, an important and complex role that should be
investigated by EC supported research efforts in the near future.

Deliverable D6
Page 15/29
Status: Final
Version: 7
Date: 30.04.08
The ETHICBOTS consortium is aware of possible positive and negative
implications for developing countries of robotic technologies and related
bionic and AI technologies. This issue has been considered here with
regard to specific applications only, for instance with regard to robotic
surgery. Additional and more comprehensive ethical research and
monitoring are clearly needed.
The ETHICBOTS consortium has not discussed in depth the issue of the
ecological impact of robotics and related bionic and AI technologies.
However, the consortium is aware that the widespread use of robots may
give rise to extensive waste disposal and recycling problems. The
consortium has not examined potential uses of robotics for
environmental protection either. Both issues are closely interrelated and
should be thoroughly examined and systematically monitored by future
EU projects.
Finally, one should take notice of the fact that the order in which topics
are listed in this report is not supposed to convey any hierarchical order
between them.

Deliverable D6
Page 16/29
Status: Final
Version: 7
Date: 30.04.08
Robotic Technologies: Ethical Monitoring and
Recommendations
Learning robots
Machine learning will play a central role in robotics for the purpose of
developing versatile service robots in general, and personal robots in
particular. There are theoretical and practical limitations in our ability to
explain, predict, and control the behaviour of learning robots in their
interactions with humans. These epistemic limitations have a significant
impact on human autonomy issues, physical integrity, moral
responsibility and liability ascription problems (see section on learning
robots in ETHICBOTS Deliverable D5).
It is recommended that in the ethical monitoring and evaluation of robotic
systems that are designed to interact with humans, one explicitly and
carefully attend to the role, if any, of machine learning techniques. In
particular, ethical monitoring is required of the potential impact of
learning technologies on moral responsibility and liability ascription
problems in service and personal robotics, in addition to physical
integrity, human autonomy and robotic system accountability issues.
Stability and Uncertainty
Robotics research and technological transfer efforts are extensively
concerned with the stability of robot sensori-motor behaviour and related
uncertainty issues (see ETHICBOTS Deliverable D5).
It is recommended that in the ethical monitoring and evaluation of robotic
systems that are designed to interact with humans, stability and
uncertainty issues be systematically and carefully attended to, assessing
their impact on moral responsibility and liability ascription problems, on
physical integrity, and on human autonomy and robotic system
accountability issues.

Deliverable D6
Page 17/29
Status: Final
Version: 7
Date: 30.04.08
Position of humans in the control hierarchy
Ethical reflection does not justify the exceptionless rule that every
individual robotic action be submitted to human supervision and approval
before its execution. This is particularly evident when human-in-the-loop
conditions jeopardize timely robotic responses, possibly leading on this
account to violation of task constraints and increased risk situations.
It is recommended that in human-robot shared action control provisions
be made for assigning humans the higher rank in the control hierarchy
which is compatible with cost-benefit and risk analyses. Furthermore, it is
recommended that robotic systems which are justifiably allowed to
override human decisions or to act independently of direct human control
or supervision be systematically evaluated from an ethical viewpoint.
Robotic action and ethically motivated design standards
Increase of legibility and predictability of robotic behaviour by humans is
a central goal of personal robotics. From an ethical viewpoint, the
promotion and protection of human dignity and autonomy, and the
related reduction of interaction risks, are powerful motivations for the
identification of design standards concerning legibility and predictability
of robot behaviour. Promotion and support of research on the
identification of such standards is recommended, involving, e.g.,
standards about recognizability of robot action-types, and the design of
robot morphological features enabling human observers to recognize the
robot as a machine and to evaluate its range of action and direction of
motion.
Privacy
IC technologies enable robots to interact remotely with any kind of
software agents hosted by other robots or computer systems. Indeed,
the broad vision driving research on converging IC and robotic systems
includes service and personal robots that are fully integrated with the
internet and the world of software agents. Accordingly, research on
converging IC and robotic technologies should be promoted and

Deliverable D6
Page 18/29
Status: Final
Version: 7
Date: 30.04.08
supported also in view of their tremendous potential for jointly extending
human action and human information and communication capabilities.
However, one should be careful to note that IC technologies enable one
to make remotely accessible “sensitive data” about humans interacting
with a robot. In particular, this information can be further processed by
any kind of connected “intelligent” softbot system which may infer and
distribute unauthorized user profiles. And clearly, this structured
information can be used to influence the behaviour of interacting humans
in illegitimate ways. Accordingly, it is recommended that any type of
robot interacting with humans be subject to a preliminary privacy impact
assessment, which includes, but is not limited to potential violations of
current privacy regulations and policies.
Further research on the possibility to use privacy enhancing technologies
in the field of robotics is also recommended.
Race, class, gender, intentional language
Suitable policies should be developed to prevent the modelling of
human-robot relationships after discriminatory or impoverished
stereotypes of, e.g., race, class, gender, personality, emotions, cognitive
capabilities, and social interaction. In particular, disregard of relevant EU
equality regulation and policies should be avoided, and one should
carefully consider the opportunity of providing suitable interpretations or
extensions of extant regulations with respect to the human-robot
interaction context.
In robotics, current uses of words such as 'knowledge', 'intelligence',
'representation', 'intention', 'emotion', 'social', 'agent', 'autonomy', and
'humanoid' are potentially misleading - insofar as it is thereby suggested
that typically human mental properties can be unproblematically
attributed to technological artefacts, disregarding in this respect current
limitations of state-of-the art robotic systems. It is recommended that
actions be taken to foster awareness among science writers, journalists
and other communication workers of the potential misuse and pitfalls of
this kind of language. Moreover, the EC may consider the possibility of

Deliverable D6
Page 19/29
Status: Final
Version: 7
Date: 30.04.08
requiring an evaluation from this psecific perspective of the language
used in proposals submitted within framework programmes and other
European level research initiatives.

Deliverable D6
Page 20/29
Status: Final
Version: 7
Date: 30.04.08
Recommendations about Domain-Specific Human-
Robot Interactions
Let us now turn to consider applications of robotic systems in specific
domains potentially involving extensive forms of human-robot interaction.
Care Robots
Care robots have the clear potential for promoting the right of people to
live a life of independence and social participation. Prospective
applications in the near future (within 5 years) are relatively limited, but
the vision driving care robot research encompasses a wide range of
applications (See ETHICBOTS Deliverable D1, p. 18. For a survey of
technologies, systems, and projects, see D1 pp. 78-93.) Research and
technological transfer in this field should be strongly encouraged and
explicitly driven by opportunities for promoting fundamental rights.
These opportunities must be pursued while protecting fundamental rights
from potential threats which arise if care robots are used to replace and
possibly to dispense altogether with human care in the assistance to
human beings in general, and to ill, disabled or elderly people in
particular. A decrease of direct human interactions induced by robotic
replacements may be highly detrimental to the well-being of assisted
people.
In accordance with this warning, the introduction of robotic care systems
should be monitored, properly justified in terms of the dignity and well-
being of their human users, and planned in view of expected
improvements in the allocation of human resources for assistance
purposes. The EC should investigate the opportunity of submitting the
deployment of care robots to the approval of ethical committees taking
proper account of the above dimensions of the problem. From a
technical point of view, research should be supported in the design and
development of care robots based on the concept of “appliance”, i.e. a
helpful device helping in everyday life but operated by a human being.

Deliverable D6
Page 21/29
Status: Final
Version: 7
Date: 30.04.08
Support of interdisciplinary research (including social sciences and
humanities) on well-being in disabled and elderly people is
recommended, which focuses on visions of assistive technologies in
general and robotic assistance in particular. These research efforts
should crucially involve potential users and human care-givers.
Technologically-informed education should be promoted in connection
with care robots, in order to improve public opinion formation and
participation in techno-ethical discussions about care robots. Similarly,
further development and support of deliberative technology assessment
procedures concerning care robots is recommended too.
Edutainment robots
Robots designed for educational purposes (say, robotic construction kits
for schools or robots acting as museum tour guides; but see, for wider
repertoires of systems and educational purposes, Deliverable D1, pp. 46-
59 and pp. 93-101) have the potential to contribute to improve education,
cultural life, creativity, and to enhance human cognitive and emotional
capabilities. Research on edutainment robots should be explicitly
encouraged and supported on these grounds, making sure that
discriminatory or impoverished stereotypes are avoided which concern,
e.g., race, class, gender, personality, emotions, cognitive capabilities,
and social interaction.
It is recommended that one treats on a par with media technology robots
that are designed and sold for entertainment purposes. Therefore,
existing regulations on media content may be applied on such robots,
both on the European level and on the level of the EU member states.
These regulations are likely to provide an effective level of protection for
minors and for human dignity.
Some robotic systems take on the external appearance of humanoid
characters or animals. Presently, these superficial analogies are not, in
general, accompanied by deeper relational, cognitive or emotional
similarities. It is recommended that psychological research on the use of
robotic toys and entertainment systems resembling humans or animals
be promoted and supported, notably with the aim of investigating and
detecting possible cognitive, emotional, and relational deficiencies
Commentato [Michael N1]: The official name of the EU
document on media regulations is: "Council Recommendation
of 24 September 1998 on the development of the
competitiveness of the European audiovisual and information
services industry by promoting national frameworks aimed at
achieving a comparable and effective level of protection of
minors and human dignity" – Therefore, I would delete "more in
general".

Deliverable D6
Page 22/29
Status: Final
Version: 7
Date: 30.04.08
arising from early age failure to categorize these robots and distinguish
them properly from animals and humans. Similarly, psychological
research should be encouraged on possible negative consequences of
deceptive simulation of emotions and the related incapability of
distinguishing between these and genuine expressions of emotions.
If the EC will further support robotic entertainment technologies and
systems, it is recommended that intended kinds of entertainment be
clearly specified and ethically evaluated.
Special caution should be used in the case of sex robots, which have
been occasionally promoted in public discussion on the medically
uncertain ground of sex therapy (let alone highly questionable claims to
the effect that robots will eventually replace human prostitutes). In case
the EC should consider financial support of research on sex robots in the
context of medical therapy, it is recommended that this be motivated on
the basis of strong medical evidence.
Sex robots with child-like appearances should be banned under any
circumstance, in accordance with the United Nations „Optional Protocol
to the Convention on the Rights of the Child on the sale of children, child
prostitution and child pornography“ (2002).
Robotic surgery
Robotic systems in surgery have the clear potential for promoting
fundamental rights, by improving the quality of medical interventions and
protecting the patient's physical integrity. In addition to this, robotic
surgery systems promise to reduce post-operatory problems, and to
allow for faster recovery.
However, in order to protect physical integrity and well-being, extensive
assessment studies are recommended concerning control which is
shared between human surgeon robot, and communication technologies
in tele-operation, by highlighting relevant responsibility and liability issues
(see the detailed case-study on ROBODOC in D5).
Thorough cost-benefit analyses are recommended, in order to assess
which benefits can flow in the near future from robotic tele-operation
Commentato [Michael N2]: I would like to focus on "Sex
robots" here, since "wargaming" is part of the regulation
pointed to above.

Deliverable D6
Page 23/29
Status: Final
Version: 7
Date: 30.04.08
technologies, especially in the way of improvements of the quality of life
of people living in developing countries which lack specialized hospital
facilities. Special attention must be devoted to problems of installation,
usage, and maintenance expense reduction which are presently hardly
affordable by developing countries. Specific research should be
supported for the porting of the expensive technologies related to
surgical robotics to low-cost technologies that are more widely
accessible.
Replacing humans in the workplace
It is to be expected that robots will increasingly replace human workers in
manufacturing processes. The ethical dimension of replacing of humans
by machines has been a key issue concerning the social impact of
robotics ever since the introduction of the first industrial robots.
On the one hand, automation technologies in general, and robotic
technologies in particular have significantly contributed to the health
quality and security of workers (for a recent development in this sense
see the robotic street cleaner case-study in D5). Arguably, the
replacement of humans by robots will serve the promotion of
fundamental rights in other ways, e. g., by contributing to reduce child
labour exploitation.
On the other hand, robotic automation in the workplace calls for ethical
evaluation, monitoring, and warning for the purpose of protecting
fundamental rights.
It is recommended that more research be promoted and supported on
the ethical implications of the impact of robots replacing humans in the
workplace. In particular, the expectation that mostly workers with low
qualification levels will continue to be replaced by machines in the near
future suggests the need for sustained research about the ethical impact
of newly emerging robotic substitutes for these forms of human labour.
Research on the impact of human-machine replacement should be
encouraged from the perspective of the robotic-divide too, since there is
hardly any systematic knowledge about the effects which the action of

Deliverable D6
Page 24/29
Status: Final
Version: 7
Date: 30.04.08
replacing humans by machines in one country brings about on the job
market of another country.
Remote control and cooperation in the workplace
Robotic systems can be tele-operated by people who are possibly
located in another country. It is recommended that existing regulations
be scrutinized to assess whether questions of responsibility and liability
for acting in tele-presence are properly taken into account. These
regulations should comprise international conventions regulating
situations in which, e.g., the human operator of a robotic system is
located in a different country as the robot providing service to someone
who is citizen of yet another country. A related question to be
investigated is whether one should be allowed to tele-operate from within
Europe a robot located in countries which do not meet European
standards for machine safety.
In situations involving mixed Human-Machine teams, it is no longer
possible to reduce human-machine-interaction to a minimum, which
seems to be at the core of existing regulations, i.e. Directive 2006/42/EC
of the European Parliament and of the Council of 17 May 2006 on
machinery. It is recommended that a new approach be developed to
address issues of safety involving interaction with new kinds of
machinery, like service and personal robots. It is also recommended that
the obligation to provide users with information on how to operate robots
expressed in existing regulations (Directive 2006/42/EC, Directive
89/391/EEC) be re-evaluated and possibly revised in light of the likely
development of more versatile and autonomous robots.
Military uses of robots
Military research in robotics is being extensively supported. Both ground
and aerial robotic systems have been deployed in warfare scenarios. It is
expected that an increasing variety and number of robotic systems will
be produced and deployed for military purposes in many developed
countries.
In view of these new military scenarios, a European level agreement
should be sought to the effect that robots are to be counted with regard

Deliverable D6
Page 25/29
Status: Final
Version: 7
Date: 30.04.08
to the Conventional Forces in Europe Treaty (CFE). More important, it is
recommended that aerial and other combat robots be included in the
preventive arms control agenda.
In view of current limitations of robotic technologies, robots do not
achieve human-level perceptual recognition performances that are
crucial, e. g., to distinguish friends or by-standers from foes. Accordingly,
tight control on and even a moratorium on autonomous combat robots
should be considered. This could be achieved by a joint effort of the EU
member states – preferably involving other OSZE countries and the
UNO. Related analyses of autonomous robotic weapon systems are
needed with regard to international warfare law, in view of the fact that
present robotic systems fail to match human performance in
discriminating between soldiers, soldiers that surrender, civilians, etc.
A detailed comparative analysis of warfare robotic technologies is
recommended with respect to control policies adopted towards other
kinds of weapons, such as the 1995 protocol banning blinding laser
weapons or the 1972 BTWC which prohibits the development of
biological weapons.
The EU member states should support research on the predictability of
behaviour and safety of aerial combat systems and consider, in the light
of results thus obtained, the possibility of restricting overflight rights on
densely populated European territory.
In view of stability, uncertainty and other kinds of unpredictability issues
concerning robotic systems at large (see D5), it is recommended that
one considers an international agreement to ban nuclear bombs and
missiles from the equipment of autonomous robotic systems.
An up-to-date definition of robot for military uses, especially in the
context of arm export, regulation and control must be provided (see also
D4, section 2.4.2, Armed Forces).
In the framework of discussion about autonomous weapon systems, the
promotion and support of interdisciplinary research is recommended on
the risks of misuse of new robotic technologies and consequences for
international security, explicitly including civil-military interaction and

Deliverable D6
Page 26/29
Status: Final
Version: 7
Date: 30.04.08
exchanges, and the operational capabilities of small groups and arms-
producing countries.
The further development of a broad ethical framework as an enabling
factor for the public to participate in discussions on dual use of robots is
highly desirable, together with deliberative technology assessment
procedures (for example consensus conferences) backed by
technologically-informed education initiatives. Suitable policies and
actions fostering awareness about the dual use of robots are highly
recommended at the level of European society. Support of extensive
initiatives in dual use problem dissemination and interdisciplinary techno-
ethics community building is recommended too.

Deliverable D6
Page 27/29
Status: Final
Version: 7
Date: 30.04.08
Connecting the human brain to robots
Bionic technologies fall in the purview of the ETHICBOTS project insofar
as these allow for the interfacing of the human brain with robotic
systems. Human-robot hybrid bionic systems have been shown to
provide effective therapeutic means to restore lost motor functionalities.
In addition to this, current bionic inquiries into human-robot hybrid
systems demonstrate a wide spectrum of possibilities for enhancing
human sensori-motor capabilities in healthy human beings. These latter
possibilities, however, are demonstrated by proof-of-concept
experimental work only. Their present or imminent interest for healthy
users is not evident in the light of cost-benefit considerations. Thus, on
the basis of the triaging dimensions adopted by the ETHICBOTS
consortium, the present section concentrates on the more imminent –
and mostly therapeutic - applications of technologies enabling one to
interface the human nervous system with robotic systems. These notably
include both invasive and non-invasive Brain-Computer Interfaces
(BCIs).
Implant technologies for connecting nervous systems to robots
In connection with invasive implants enabling connections between
nervous systems and robotic systems (see specific sections in D1 and
D5), the ETHICBOTS consortium endorses the ethical analyses provided
in the above mentioned opinion on ICT implants in the human body by
the European Group on Ethics in Science and New Technologies to the
European Commission: ‘Ethical aspects of ICT implants in the human
body’, Opinion No. 20, Adopted on 16/03/2005. However, it is
recommended that the EGE Opinion be regarded just as a starting point
for further ethical research, which is clearly needed in view of the fact
that this research is rapidly evolving and its results directly affect the
brain and correlated mental capabilities. Ethical research should be
promoted and supported on ICT implants in general, and implants
interfacing the nervous system with robotic systems in particular.

Deliverable D6
Page 28/29
Status: Final
Version: 7
Date: 30.04.08
Non-invasive BCIs and robotic system control
Non-invasive BCIs involve no implant technology and are mostly used to
control computer or robotic devices in therapeutic applications enabling
one to restore lost communication and motor capabilities (see specific
D5 section on such devices). BCIs have been shown to provide effective
means to restore lost communication and motor capabilities in patients
paralysed by spinal chord injuries or muscular dystrophies, thereby
helping severely disabled people to increase their independence and to
participate in social life. Brain-actuated devices presently include virtual
computer keyboards, robotic wheelchairs, and robotic manipulators.
The concept of direct brain-robot communication emerging from the
recent progresses made in this young field of research is to build on
these initial results and technologies in order to extend communication
and control capabilities of both disabled and non-disabled individuals. It
is recommended that extensive ethical research be promoted and
supported concerning both the imminent therapeutic applications of BCI-
robot integration technologies and the emerging promise of BCI-robot
integration. In particular, additional ethical research is needed along the
following chief dimensions.
Autonomy
BCI-robot integration has a clear potential for promoting autonomy in
people affected by severe motor disabilities. However, a trade-off arises
in view of the need of relying on shared control of action due to the
relatively limited range of action commands that present-day BCI
technologies enable one to communicate to robots. Further research on
the promotion and protection of personal autonomy is recommended in
connection with this particular trade-off and the possibility of
misinterpreting user commands depending on brain plasticity and
limitations of machine learning methods.
Responsibility
Misinterpretations of user intents and commands expressed by means of
a BCI may depend on theoretical and practical limitations of sensor
technology and machine learning methods in computer science. More

Deliverable D6
Page 29/29
Status: Final
Version: 7
Date: 30.04.08
research is needed in order to identify and analyse moral responsibility
and liability ascription issues arising from these technological limitations
in the context of robotic systems whose control policy involves BCI input
commands.
Distributive justice
Further research is recommended on distributive justice issues arising in
connection with non-discriminatory access to therapeutic BCI-enabled
robotic systems and the distribution of limited healthcare resources.
Special attention should be paid to an analysis of preventive medicine
uses of BCIs to preserve sensory motor and cognitive abilities in
degenerating diseases leading to locked-in states.
Privacy
More research on privacy protection issues is recommended in
connection with mental and brain state reading possibilities afforded by
BCI research, and the broadcasting of information about brain and
mental states which is made possible by computers and robots that are
connected to BCIs. This recommendation extends the releated
recommendations on privacy and robots made in the previous section.
Personhood
Further research on the protection of personal integrity is recommended
along the dimensions of personality changes, personal identity
persistence, and social identity and identification problems which may
arise from both invasive and non-invasive BCI-enabled connections with
robotic systems. In particular, ethical problems related to the narrative
dimension of personal identity, as well as the bodily and mental
continuity criteria for personal identity stand in need of more extensive
investigation, which significantly bear on legal issues concerning
autonomy, liability and responsibility arising in this technological context
of human-machine inetraction.