Towards a Theory of Awareness
Ricardo Sanz
1 a
, Manuel Rodr
´
ıguez
1 b
, Mart
´
ın Molina
2 c
, Esther Aguado
1 d
and Virgilio G
´
omez
1 e
1
Autonomous Systems Laboratory, Universidad Polit
´
ecnica de Madrid,
c/ Jos
´
e Gutierrez. Abascal 2, 28006 Madrid, Spain
2
Artficial Intelligence Department, Universidad Polit
´
ecnica de Madrid,
Campus de Montegancedo, 28660 Boadilla del Monte, Madrid, Spain
Keywords:
Awareness, Theory, Autonomy, Cognitive Science, System Architecture, Engineering Methods, Science,
Consciousness, Perception, Attention, Understanding.
Abstract:
Observation of humans and animals shows that awareness is a critical aspect of mental processes for those
agents that operate in changing environments. Responding to potentially dangerous situations and leverag-
ing environmental affordances are essential capabilities for autonomous agents’ ecological viability. Agents
need to be aware of their situations. Artificial autonomous systems construction depend on using suitable
system architectures and applying proven engineering methods. While current systems display a certain de-
gree of awareness, it is unclear what principles shall be used in their design. We are in a pre-scientific,
pre-technological situation concering awareness. Unfortunately, the scientific analysis of the awareness phe-
nomena is quite difficult because its principles cannot be easily isolated in fully functioning human minds. We
need a clean, formal theory of general awareness of universal nature. This theory should be applicable both
to humans and machines, and not exclusively bound to the psychology and neurobiology of living animals. In
this position paper, the authors argue for developing such a theory, state some requirements for it and propose
an initial conceptual seed for a future theory of awareness that orbit around the idea that awareness is the
real-time understanding of sensory flows.
1 INTRODUCTION
The mere observation of humans’ and animals’ be-
haviours shows that awareness is a central aspect of
mentality for those agents that successfully operate
in changing, challenging environments. Perceiving
change, responding to potentially dangerous situa-
tions and leveraging environmental affordances are
essential capabilities for autonomous agents ecolog-
ical viability, i.e. agents that are capable of surviv-
ing the disturbances that a non-controlled environ-
ment throw on them. So should be for robots and
other classes of situated machines. Endowing ma-
chines with “awareness” should improve their auton-
omy profile, making them more reactive to world dy-
namics (Sanz et al., 2007a). However, this is not an
a
https://orcid.org/0000-0002-2381-933X
b
https://orcid.org/0000-0003-0929-5477
c
https://orcid.org/0000-0001-7145-1974
d
https://orcid.org/0000-0002-7860-9030
e
https://orcid.org/0000-0001-8538-5111
easy task; especially because we do not have a good,
translatable Theory of Awareness (ToA).
In this domain, authors sometimes use the term
“awareness” and sometimes use the term “conscious-
ness”. The use of the two terms has similarities and
differences that may vary across disciplines, domains
and languages. It is not easy to pinpoint the differ-
ence between “consciousness” and “awareness”, as
Francis Crick acknowledged (Crick and Koch, 1992).
From some perspectives, we can consider both the
same thing, e.g. in the perceptual domain; from oth-
ers, we cannot, e.g. in the ethical domain. We will
follow here Crick’s policy of considering them syn-
onyms; giving preference to “awareness” and using
“consciousness” only when trying to make a specific
point concerning the term. However, other authors do
follow other policies and use both terms (as the refer-
ences and theories will show).
In principle, it should be feasible to translate most
human mental traits into machine mental traits thanks
to the general multiple realizability of physical sys-
Sanz, R., Rodríguez, M., Molina, M., Aguado, E. and Gómez, V.
Towards a Theory of Awareness.
DOI: 10.5220/0012595100003636
Paper published under CC license (CC BY-NC-ND 4.0)
In Proceedings of the 16th International Conference on Agents and Artificial Intelligence (ICAART 2024) - Volume 3, pages 1413-1420
ISBN: 978-989-758-680-4; ISSN: 2184-433X
Proceedings Copyright © 2024 by SCITEPRESS Science and Technology Publications, Lda.
1413
tems
1
. To translate awareness from humans to ma-
chines we only need a physical theory of awareness.
However, in the case of awareness-related aspects,
there is a strong risk of falling into a very deep rabbit
hole of biologism if we are not careful enough as to
be perfectly clear on what we are talking about and
what class of theory do we need. In this paper we
argue for the development of such a general theory
of awareness, trying to be careful enough as to avoid
the rabbit hole, and delivering a theory applicable to
animals and translatable to machines.
2 AWARENESS IN ROBOTS
The topics of consciousness and awareness have been
quite marginal in the domains of AI and robotics
(Chella and Manzotti, 2007; Chella, 2023).
The efforts to build conscious AIs or the discus-
sion about its very possibility was a matter of a small
group of researchers with some philosophical voca-
tion. Only recently this community has gained some
extra human mass, esp. when some comments by
well-known actors in the current AI landscape hit the
media. These comments talked about the possibility
of current mainstream AI implementations could be
reaching a state of consciousness. These reminded
people about Skynet becoming aware and trying to
kill humanity in Cameron’s classic Terminator. How-
ever, even when this is becoming a wider discussion,
the research program on aware AI is still very flaky;
but it should be not if we are right concerning its im-
portance for autonomous machines.
In the last decades, some researchers have at-
tempted the creation of real implementations of AIs
and robots with consciousness. In most cases, the
approach take was 1) select one of more char-
acteristics associated to human consciousness, and
2) develop a machine that demonstrated this char-
acteristic. For example, Aleksander addressed
anticipation and imagination (Aleksander, 2009),
Tani addressed mirror self recognition (Tani, 2017),
Chella addressed inner speech (Chella et al., 2020),
Hoffmann addressed proprioceptive self-awareness
(Hoffmann, 2021), Hern
´
andez addressed metacog-
nitive self-awareness (Hern
´
andez et al., 2009), and,
brave enough, Haikonen addressed qualia (Haikonen,
2013).
The attempts will continue, for example extending
into the very active domains of large, language-based
AIs. However, the current statistical approaches of
machine learning from human-bound data will not
1
We will here take a strict, non-dualist, physicalist
stance concerning mental aspects like awareness.
achieve the desired end of powering the engineering
capability of creating custom awareness
2
. The en-
gineering of these systems will only be effective if
grounded in more profound, structural theories that
seem far from what current machine learning capa-
bility can provide. As Dacey says, (Dacey, 2022)
“Statistical inference cannot do all of the work of the-
ory choice. We have the need of producing a proper
structural deep theory of awareness to do both science
and technology.
3 RATIONALE FOR A THEORY
OF AWARENESS
The rationale for seeking a theory of awareness has
deep grounds in science, philosophy and engineering
(Sanz et al., 2007a).
Understanding awareness needs an approach from
a unified perspective. For example, in the realm
of philosophy, awareness is related to four essential
branches: epistemology -what the agent gets to know
through the senses, ontology -what the agent is aware
of-, phenomenology -the sensations that the agent
gets-, axiology -the value that such perceptions have
for the agent dwellings. Awareness seems essential
for dwelling in open dynamic worlds.
This need is also very relevant in the world of arti-
ficial systems. Engineers are seeking system design
solutions to deal with the uncertainty of the world
and the uncertainty of the systems themselves. In
many cases, autonomous robots failures are not due
to changes in the world but changes in the robot it-
self (e.g. failures or emergent phenomena in the
robot software subsystems). The streamlined con-
struction and dependable runtime operation of arti-
ficial autonomous systems depend on using suitable
system architectures and applying proven engineering
methods (Aguado et al., 2021). Having solid theo-
ries of world-awareness and self-awareness may help
achieve the desired results.
In this sense, a general theory of awareness is a de-
sirable asset for both scientists and engineers. Unfor-
tunately, the scientific analysis of the awareness phe-
nomena is quite difficult because it cannot be easily
identified and isolated in fully functioning biological
minds. We need a single theory of general aware-
ness of universal nature and this universality seems
difficult, especially when we seek a theory applicable
both to humans and robots, and not exclusively bound
2
Awareness that is designed and scaled to the needs of
the target technical system that need not be similar to a hu-
man.
AWAI 2024 - Special Session on AI with Awareness Inside
1414
to the psychology and neurobiology of living animals
(Wilson, 1998). Achieving an universal –applicable
to all classes of systems– and unified –explaining all
related phenomena— theory is a complex challenge.
Besides the different classes of subjects, there are too
many aspects of consciousness to deal with. Alek-
sander identified five aspects of consciousness to be
mapped into machines —perception, imagination, at-
tention, planning, emotion— but, for example, Tani
identified ten different aspects —based on the phe-
nomenological analysis of Husserl.
Long ago, Aaron Sloman said that “It is not worth
asking how to define consciousness, how to explain
it, how it evolved, what its function is, etc., because
there’s no one thing for which all the answers would
be the same. Instead, we have many sub-capabilities,
for which the answers are different: e.g. different
kinds of perception, learning, knowledge, attention
control, self-monitoring, self-control, etc.
3
Many authors, especially in the biological and hu-
manities domains, argue that machine consciousness
is impossible. However, the possibility of devising a
single mechanism explaining all the aspects of aware-
ness in a system-neutral sense, or at least separat-
ing them into related and unrelated aspect as Sloman
proposed, need not be an impossible dream (Hadley,
2023). As Francis Crick said:
The second assumption is tentative: that
all the different aspects of consciousness, for
example pain and visual awareness, employ a
basic common mechanism or perhaps a few
such mechanisms. If we understand the mech-
anisms for one aspect, we will have gone most
of the way to understanding them all. (Crick
and Koch, 1990)
In this position paper, the authors argue for de-
ploying effort towards developing such a theory, stat-
ing some requirements for it and proposing an initial
seed for a potential future Theory of Awareness.
4 THEORIES OF AWARENESS
There are many theories that offer different perspec-
tives on the nature of awareness, reflecting the inter-
disciplinary nature of the field. See for example the
article by Seth and Bayne (Seth and Bayne, 2022) for
a more exhaustive review in the domain of biological
consciousness.
3
In comp.ai.philosophy, 14 Dec. 1994.
4.1 Perspectives on Awareness
Here are some important theories and perspectives re-
lated to awareness:
In the domain of cognitive psychology:
Selective Attention: Awareness seems closely
related to selective attention. We are con-
sciously aware of the information we selec-
tively attend to, and other stimuli may not enter
our conscious awareness (Taylor, 2002).
Levels of Processing: Depth of processing in-
formation affects awareness (Craik and Lock-
hart, 1972). Deeper elaboration leads to better
retention and awareness of information.
In the domain of neuroscience:
Global Workspace Theory (GWT): GWT
(Baars, 1997) says that conscious awareness
arises from the global broadcast of information
throughout the brain. Certain neural processes
involve a global workspace that integrates in-
formation and brings it to conscious awareness.
Neural Correlates of Consciousness (NCC):
Researchers seek to identify specific neural ac-
tivity patterns associated with conscious aware-
ness (Koch et al., 2016). Understanding these
neural correlates can shed light on how the
brain generates awareness.
In the domain of Philosophy:
Phenomenal Consciousness: This perspective
explores the nature of subjective experience —
having “qualia. It analyses what it is like to
have a particular experience and how subjective
awareness happens (Nagel, 1974).
Higher-Order Thought (HOT) Theory: This
theory proposes that awareness arises from the
ability to have thoughts about one’s own mental
states (Rosenthal, 2000).
In the domain of social psychology:
Social Awareness: How individuals are aware
of and respond to the social environment
(Durkheim, 1893). It includes theories related
to social perception, empathy, and understand-
ing the mental states of others.
In the domain of general systems:
Integrated Information Theory (IIT): IIT
(Tononi, 2004) says that a system is conscious
if it has a high degree of integrated information.
This theory has received continuous support
thanks to its formal nature.
In the domain of quantum mechanics:
Towards a Theory of Awareness
1415
Orchestrated Objective Reduction (Orch-OR):
Orch-OR (Hameroff and Penrose, 2014) sug-
gests that quantum processes in microtubules
within brain cells play a role in consciousness.
Besides the effort put and the raising interest in the
topic, the study of awareness is still in a pre-scientific
stage. There are lots of ongoing research and debate,
but no single theory has gained unanimous acceptance
(Jylkk
¨
a and Railo, 2019). See, for example, the re-
cent polemics about IIT being considered as pseudo-
science by some authors, esp. in the neuropsychology
domain (Fleming, 2023).
4.2 Awareness vs Consciousness
As said at the beginning, the terms “awareness”
and “consciousness” are often used interchangeably.
However the two terms are sometimes used differ-
ently in various contexts, and there isn’t a universally
agreed-upon distinction.
Different perspectives —GWT, IIT, Orch-OR,
etc.— attempt to differentiate awareness from con-
sciousness but these distinctions are not universally
accepted. Different theories and disciplines may use
the terms in varying ways and explain them using con-
ceptualizations that are far from beign harmonised.
The field remains dynamic, and our understanding
of consciousness and awareness is likely to evolve
with ongoing research and interdisciplinary explo-
ration and the elaboration of proper, transversal the-
ories.
A distinction that may be somewhat distilled from
the previous list is that awareness is related to infor-
mation –i.e. to the epistemic aspect– and conscious-
ness to sentience –i.e. to the phenomenic aspect. The
work that we are doing points into the epistemic direc-
tion, leaving the sentience aspect to ulterior scientific
efforts.
5 THE CONTENT OF A THEORY
The ToA shall be a theory that is both scientific and
operationalisable. As a scientific theory, the ToA shall
be a well-substantiated explanation of some aspect of
the natural world that is based on a body of evidence,
observations, and experiments. This body of evidence
comes from the cognitive operation of animals and
also from the cognitive operation of machines, esp. in
the uncertain domains of the open robot world. The
“aspect of the natural world” that we are interested in
is the phenomenon of “awareness”.
We want the ToA to be a solid theory —a robust
and well-established scientific explanation— to serve
as the framework upon which scientists could build
their understanding of the phenomenon of awareness
and pile-up solid research results, and engineers use
this understanding in its operationalisation as applied
science in building better cognitive robots.
5.1 Characteristics of a Scientific
Theory
This is a list of key components and characteristics of
a scientific theory and to what extent the ToA shall
address them:
Empirical Basis: A scientific theory is grounded in
empirical evidence. It should be supported by
a substantial body of observations, experiments,
and data collected through systematic and repeat-
able methods. This empirical support is crucial in
distinguishing a theory from a mere hypothesis or
conjecture. The source of evidence is deployed
human cognition and the situations where aware-
ness plays a central role. For example, the META-
TOOL project
4
explores evidence concerning the
role of awareness in ancient tool making. Obvi-
ously we are dealing with cognitively problematic
situations (Norman, 1980) and hence, a proper or-
ganization of evidence will be critical.
Consistency: A scientific theory must be internally
consistent, meaning its various components and
principles should not contradict one another. It
should provide a logical and coherent framework
for explaining observed phenomena. It shall
also be consistent with other accepted scientific
theories. The use of formal methods –as IIT
attempted– and the model-based methods of en-
gineering may provide the necessary support to
guarantee this consistence.
Testability: Scientific theories are falsifiable (Pop-
per, 1959). They can be subjected to experimen-
tation and observation, and there should be clear
criteria that, if not met, would disprove the the-
ory. The ability to test and potentially disprove
a theory is a fundamental aspect of the scientific
method. The level of robot performance in real
settings may provide this necessary evidence.
Predictive Power: A strong scientific theory can
make predictions about future observations or ex-
periments. These predictions should be based on
the theory’s principles and should be verifiable
through empirical testing. The theory’s ability to
make accurate predictions lends further credibility
to it. This is an essential aspect for a theory that is
4
http://metatool-project.eu
AWAI 2024 - Special Session on AI with Awareness Inside
1416
used as a design asset in an engineering endeav-
our. Engineers will use designs based on this the-
ory to guarantee effectiveness in future systems.
Scope and Explanatory Power: A scientific theory
should have a broad scope, meaning it can explain
a wide range of related phenomena. The more
phenomena it can explain, the more powerful and
influential the theory is. This sits at the very core
of the scientist ambition: we target a theory that
not only addresses robotic awareness but general
cognitive systems awareness (see Figure 1).
Figure 1: Target a general cognitive theories of awareness.
Simplicity: If two or more theories explain the same
phenomena equally well, the simpler one is to be
preferred (Ockam’s razor). Simplicity makes the-
ories more elegant and easier to work with. So far,
there is no real competition in theories of aware-
ness (i.e. in the terms stated here). We shall try to
make the ToA simple using compact and effective
abstractions. The CORESENSE project
5
tries to
use category theory for these abstractions in part
motivated by this search for simplicity.
Reproducibility: The experiments should be repro-
ducible by other scientists using the same meth-
ods and conditions. This is a cornerstone of the
scientific method and ensures the reliability of the
theory. This has always been a problem in cog-
nitive robotics and in cognitive systems in gen-
eral. To this end, benchmarking has been used
in robotics to enhance this reproducibility. For
example, the RoboCup@Home challenge used in
robotics specifically addresses this problem.
Peer Review: Experts in the field assess the theory’s
validity and the quality of the evidence support-
ing it. Peer review is an important process for
maintaining the rigor and reliability of scientific
theories. The Open Science approach of modern
research specifically addresses this need.
5
http://coresense.eu
Consensus: While scientific consensus can evolve
and change, a well-established theory is gener-
ally widely accepted within the scientific commu-
nity (achieving the status of “normal science” in
Kuhn terms or becoming a “framework of shared
commitments” (Eckardt, 1995)). In the domain of
psychology this is not easy. Psychological con-
structs suffer from the “toothbrush” problem: no
self-respecting psychologist wants to use anyone
else’s. But in rigorous science, we are sometimes
forced to do so by the force of the facts. If our the-
ory is solid, well documented and effective, con-
sensus will eventually emerge.
Understandability: As quantum mechanics demon-
strate, understandability is not a necessary char-
acteristic of scientific theory. However, we would
like our theory to be understandable by a broad
community of scientists and engineers. This may
require from us the expression of the theory in dif-
ferent ways that can reach these people.
In science, alive theories are not absolute truths
but are just our best current explanations based on the
available evidence. The process of forming and refin-
ing scientific theories is the essential ongoing and dy-
namic aspect of the scientific endeavour. In a sense,
this filtering-out by evidence can only happen when
the abstract concepts are strictly mapped into more
concrete realities. Most of the discussion on aware-
ness is pre-scientific in this sense: deals with abstrac-
tions disconnected from the evidence.
5.2 Operationalisation of a Scientific
Theory
The term “operationalizable theory” is not a standard
concept in scientific terminology. It is a combina-
tion of two important concepts: “theory” and “opera-
tionalization.
Theory: As described before, a theory is a well-
substantiated explanation of some aspect of the
natural world that is based on empirical evidence
and can be used to make predictions and under-
stand phenomena.
Operationalization: Operationalization is a process
in research where abstract concepts or variables
are defined and measured in a concrete and ob-
servable way. It involves specifying how a the-
oretical concept will be measured or observed in
practice. This is a crucial step in turning abstract
theories into testable hypotheses and conducting
empirical research.
The term “operationalizable theory” describes a
theory that has been sufficiently developed and de-
Towards a Theory of Awareness
1417
fined so that its key concepts and variables can be op-
erationalized for empirical experimentation. In this
context, an operationalizable theory would be one that
can be translated into specific, measurable variables
or constructs that researchers can work with to con-
duct experiments, gather data, and test hypotheses
(e.g. building AI-driven robots and deploying them in
open worlds). This operationalization is a crucial step
in the scientific method when examining the applica-
bility and validity of a theory in real-world scenarios.
6 INITIAL STEPS TO A THEORY
OF AWARENESS
In the CORESENSE and METATOOL projects we
have the specific task of developing a ToA. Any the-
ory is a complex construct that consists of several
key elements. These elements shall work together to
provide a comprehensive and well-substantiated view
of some aspect of the worlds. In some sense this
view is explanatory e.g. when applied to humans
or animals— and in another sense this view is oper-
ational —as when it enables the construction of arte-
facts.
6.1 Essential Elements for a Theory
Table 1 describes some specific elements and aspects
that a theory typically has. It is too early in the de-
velopment of our ToA to detail how the theory ad-
dresses all of them. In this paper we will just address
the domain and the initial concepts. The domain is,
obviously, “awareness”. The initial concepts are sum-
marily described in Section 6.2.
The elements and structure of a theory may vary
depending on the field of science and the nature of
the phenomenon being studied. This may imply some
specifics of the ToA when applied to certain classes
of systems. However, these elements collectively
contribute to the development of a robust and well-
supported scientific theory.
6.2 Initial Concepts for a Theory of
Awareness
When considering the domain of the theory, we shall
be aware that both consciousness and awareness are
mongrel concepts when applied to humans: They are
used in many senses, referring to different classes of
phenomena, generating confusion and long irrelevant
discussions.
An analytical effort is necessary to separate the
different mental aspects they refer to and a termino-
logical effort will be necessary to suitably label all
those aspects. In some cases, the terminological effort
is addressed by using noun phrases like “visual aware-
ness” or “synthetic awareness”, but this usually im-
plies a subclassing from a general “awareness” class.
Another approach is the use of prefixes to create new
terms when the intention is to create terms for unre-
lated classes of phenomena. Examples of this are P-
consciousness and A-consciousness for phenomenal
consciousness and access consciousness.
In this paper the terms are used in a very specific
sense: we are interested in the functional aspects of
awareness that let an autonomous system act prop-
erly when it is aware of the situation it is immersed
in. In our case, the situational awareness proposed by
Endsley (Endsley, 1995) address not only the system
environment but extends to 1) the system itself and 2)
its relation to the environment when pursuing an ex-
ternally imposed mission. The fundamental domain
entities in this scenario are: the autonomous system is
the subject of awareness, generating a mental model
of some objects situated in the environment, the part
of the world that is causally connected with both sub-
ject and objects.
In this base scenario of a subject being aware of an
object, we investigate the essential character of aware-
ness processes and how they are related to perception
and understanding. We are interested in the processes
that underlie the capability of the subject to be aware
and understand the changes in the object to be more
effective in completing its mission.
As part of these initial steps, these are some fun-
damental concepts under elaboration in this research:
Sensing: Getting information –sense-data– bound to
an object in the environment.
Perceiving: Integration of the sensory information
into a model of the object by means of a modelet.
Modelet: A partial model related to a target system.
A information structure that sustains a modelling
relation (Rosen, 2012).
Model: Integrated actionable representation; an inte-
grated set of modelets.
Engine: Set of operations over a model (e.g., integra-
tion of a modelet, exertion, compaction, pruning,
intensification, chunking, etc).
Inference: Derive conclusions from the model.
Valid inference: A inference whose result matches
the phenomenon at the modelled object.
Exert a model: Perform valid inferences from the
model.
AWAI 2024 - Special Session on AI with Awareness Inside
1418
Table 1: Essential Elements and Aspects of a Theory.
Element
Content of the element
Phenomenon
or domain
A theory explains a specific natural phenomenon or a particular domain of inquiry. It de-
fines the scope of what it seeks to explain. In our case: “awareness”.
Concepts The theory contains a set of well-defined concepts that provide the vocabulary and frame-
work for discussing and understanding the phenomenon.
Hypotheses The theory often generates specific hypotheses, i.e. testable predictions about how certain
variables or factors are related within the defined domain, and guide empirical research.
Laws or Prin-
ciples
A theory may incorporate laws or principles typically derived from empirical data and ob-
servations, that describe relationships or patterns observed within the phenomenon.
Relationships The theory specifies causal relationships between the concepts and variables involved.
Explanatory
Power
A theory should have a high degree of explanatory power, meaning it can account for a wide
range of observations and data within its domain.
Predictive
Power
A strong theory can make accurate predictions about future observations or experiments, i.e.
should be verifiable through empirical testing.
Models In some scientific theories, especially in the physical sciences and engineering, formal mod-
els may be used to describe and predict the behaviour of the phenomenon.
Empirical
Support
A theory is grounded in empirical evidence. It should be supported by a substantial body of
observations, experiments, and data collected through systematic and repeatable methods.
Evolution Scientific theories are subject to revision as new evidence and understanding emerge.
Consistency A scientific theory must be internally consistent, meaning its various components and prin-
ciples should not contradict each other.
Reviews Scientific theories are typically subjected to peer review, where experts in the field assess
the theory’s validity and the quality of the evidence supporting it.
Understanding: Achieving exertability of a modelet;
e.g. by mental model integration of a modelet and
activation of associated engines.
Specific understanding: Understanding concerning
a specific set of exertions (can be extensive or in-
tensive).
Structure understanding: Understanding the struc-
ture of the object implies achieving exertability
concerning the system structure.
Behaviour understanding: Understanding the be-
haviour of the object (achieving exertability con-
cerning the system behaviour).
Mission understanding: Understanding mission-
bound exertions (i.e. achieving derivability of
valid results from the model that can be used by
the agent to fulfill the mission).
Awareness: Real-time understanding of sensory
flows.
Awareness of : Object-bound awareness.
Self-awareness: Subject-bound awareness. Aware-
ness concerning inner perceptive flows.
The current project work is related to the devel-
opment of the formal model of the theory of aware-
ness in the form of 1) a formal ontology in higher-
order logic and 2) an architecture expressed in formal
MBSE
6
languages.
7 CONCLUSIONS
The elaboration of formal theories of awareness is ac-
tive (Bringsjord and Sundar, 2020) but still lost in
the variety of phenomena associated to conscious-
ness. An analytical approach at clarifying the many
aspects of mentality and awareness is necessary. This
means properly setting the boundaries of the domain
of explanation and being more precise on the class of
phenomena that the theory is addressing.
In this paper we have described the overall ex-
pected content of a Theory of Awareness to be of
applicability in the construction of broad classes of
autonomous systems. The elaboration of the causal
principles and laws will enable the development of ar-
chitectural patterns that will enable the design of new
systems (Sanz et al., 2007b). Besides architectural
patterns, this theory could also provide many other
resources in autonomous systems engineering to an-
alyze existing systems, to guide new designs, or to
build specific programs, etc. (examples of specific re-
6
MBSE is an acronym of Model-Based Systems Engi-
neering, a way of performing systems engineering where
models are the central assets.
Towards a Theory of Awareness
1419
sources are design idioms, complete reference archi-
tectures, common terminologies formalized through
ontologies, domain specific languages, reusable com-
ponents in software libraries, etc.).
See more about these developments at the CORE-
SENSE
7
and METATOOL
8
project websites and the
Awareness Inside
9
EIC Pathfinder Challenge.
ACKNOWLEDGMENTS
The authors acknowledge the support of the European
Commission through the Horizon Europe project
#101070254 CORESENSE and the European Inno-
vation Council project #101070940 METATOOL.
REFERENCES
Aguado, E., Milosevic, Z., Hern
´
andez, C., Sanz, R., Gar-
zon, M., Bozhinoski, D., and Rossi, C. (2021). Func-
tional self-awareness and metacontrol for underwater
robot autonomy. Sensors, 21(4):1–28.
Aleksander, I. (2009). Designing Conscious Systems. Cog-
nitive Computation, 1(1):22–28.
Baars, B. J. (1997). In the theater of consciousness: the
workspace of the mind. Oxford University Press, New
York.
Bringsjord, S. and Sundar, G. (2020). The Theory of Cog-
nitive Consciousness, and Λ (Lambda). Journal of Ar-
tificial Intelligence and Consciousness, 07:1–27.
Chella, A., editor (2023). Computational Approaches To
Conscious Artificial Intelligence. World Scientific.
Chella, A. and Manzotti, R., editors (2007). Artificial Con-
sciousness. Imprint Academic.
Chella, A., Pipitone, A., Morin, A., and Racy, F. (2020). De-
veloping Self-Awareness in Robots via Inner Speech.
Frontiers in Robotics and AI, 7:16.
Craik, F. I. and Lockhart, R. S. (1972). Levels of processing:
A framework for memory research. Journal of Verbal
Learning and Verbal Behavior, 11(6):671–684.
Crick, F. and Koch, C. (1992). The Problem of Conscious-
ness. Scientific American, 267(3):152–159.
Crick, F. and Koch, C. C. (1990). Towards a neurobiolog-
ical theory of consciousness. Seminars in the Neuro-
sciences, 2:263–275.
Dacey, M. (2022). Separate substantive from statistical hy-
potheses and treat them differently. Behavioral and
Brain Sciences, 45:e9.
Durkheim,
´
E. (1893). De la division du travail social.
Presses Universitaires de France.
Eckardt, B. V. (1995). What is Cognitive Science? Bradford
Books. The MIT Press.
7
http://coresense.eu
8
http://metatool-project.eu
9
https://awarenessinside.eu/
Endsley, M. R. (1995). Toward a theory of situation aware-
ness in dynamic systems. Human Factors, 37(1):32–
64.
Fleming, S. M. e. a. (2023). The Integration Information
Theory As Pseudoscience.
Hadley, M. J. (2023). A Generic Model of Consciousness.
Journal of Artificial Intelligence and Consciousness,
10(02):291–308.
Haikonen, P. O. A. (2013). Consciousness and the Quest
for Sentient Robots. In Chella, A., Pirrone, R., Sor-
bello, R., and J
´
ohannsd
´
ottir, K., editors, Biologically
Inspired Cognitive Architectures 2012, pages 19–27.
Springer.
Hameroff, S. and Penrose, R. (2014). Consciousness in the
universe. Physics of Life Reviews, 11(1):39–78.
Hern
´
andez, C., L
´
opez, I., and Sanz, R. (2009). The Opera-
tive Mind: a Functional, Computational and Modeling
Approach To Machine Consciousness. International
Journal of Machine Consciousness, 1(1):83–96.
Hoffmann, M. e. a. (2021). Robot in the Mirror: Toward
an Embodied Computational Model of Mirror Self-
Recognition. KI - Kunstliche Intelligenz, 35(1):37–51.
Jylkk
¨
a, J. and Railo, H. (2019). Consciousness as a concrete
physical phenomenon. Consciousness and Cognition,
74:102779.
Koch, C., Massimini, M., Boly, M., and Tononi, G. (2016).
Neural correlates of consciousness: progress and
problems. Nat Rev Neurosci, 17(5):307–321.
Nagel, T. (1974). What Is It Like to Be a Bat? The Philo-
sophical Review, 83(4):435–450.
Norman, D. A. (1980). Twelve Issues for Cognitive Sci-
ence. Cognitive Science, 4:1–32.
Popper, K. (1959). The Logic of Scientific Discovery. Basic
Books.
Rosen, R. (2012). Anticipatory Systems. Philosophical,
Mathematical, and Methodological Foundations, vol-
ume 1 of IFSR International Series on Systems Sci-
ence and Engineering. Springer, 2nd edition.
Rosenthal, D. M. (2000). Metacognition and Higher-Order
Thoughts. Consciousness and Cognition, 9:231–242.
Sanz, R., L
´
opez, I., and Bermejo-Alonso, J. (2007a). A
Rationale and Vision for Machine Consciousness in
Complex Controllers. In Chella, A. and Manzotti, R.,
editors, Artificial Consciousness. Imprint Academic.
Sanz, R., L
´
opez, I., Rodr
´
ıguez, M., and Hern
´
andez, C.
(2007b). Principles for consciousness in integrated
cognitive control. Neural Networks, 20(9):938–946.
Seth, A. K. and Bayne, T. (2022). Theories of conscious-
ness. Nature Reviews Neuroscience, 23(7).
Tani, J. (2017). Exploring robotic minds : actions, sym-
bols, and consciousness as self-organizing dynamic
phenomena. Oxford University Press.
Taylor, J. G. (2002). How Can the Self Understand Itself?
A Review of Models of the Self Edited by Shaun Gal-
lagher and Jonathan Shear. Psyche, 8(12).
Tononi, G. (2004). An information integration theory of
consciousness. BMC Neuroscience, 5(42).
Wilson, E. O. (1998). Consilience. The Unity of Knowledge.
Alfred A. Knopf, New York, USA.
AWAI 2024 - Special Session on AI with Awareness Inside
1420