Quantum-Informed Cybernetics for Collective Intelligence in IoT Systems

Featured Application

We propose an extended cybernetic model that incorporates a quantum-informed framework that maps collective systems

Abstract

Collective intelligence within a quantum-informed cybernetic paradigm presents a transformative perspective to examine adaptability and resilience in Internet of Things (IoT) systems. This paper introduces Cogitor5, a fifth-order cybernetic system that builds upon the foundational principles of the fourth-order COgITOR framework, a liquid computational system designed for complex adaptive processes. The term COgITOR is etymologically linked to the Latin passive verb cogĭtur, translating to “He is gathered,” in contrast to the more commonly recognized active form cogito, meaning “I gather” or “I think,” as famously articulated by Descartes. In contrast to conventional binary systems, Cogitor5 functions as a simulation-based complex adaptive system, inspired by a population of nano agents represented by nanoparticles suspended in a colloidal medium. These agents exhibit autonomous interactions within the solvent, featuring quantum-enabled properties that facilitate advanced self-organization and coevolutionary dynamics. This intricate model captures the complexities of agent interaction, offering a refined representation of their evolving collective intelligence. The study redefines collective intelligence as emergent process intelligence, relevant to the adaptive capacities of both biological and cybernetic systems. By utilizing metacybernetic principles in conjunction with theories of complex adaptive systems, this paper investigates how IoT networks can evolve to enhance agency trajectory formation and increase adaptability. Cogitor5 serves as an innovative computational framework for addressing the inherent complexities of IoT, providing clarity in examining self-organization, self-regulation, self-maintenance, and sustainability, thus elevating system viability. The methodology encompasses the modeling of collective and process intelligence within the scope of Mindset Agency Theory (MAT), an advanced metacybernetic model that allows for evaluable characteristics. Furthermore, this approach integrates theoretical modelling and a practical case study implemented in Matlab^® to illustrate agency functionality within a dynamic system simulating failures in the nodes of an electric grid.

1. Introduction

The Internet of Things (IoT) has evolved from isolated devices into interconnected systems enhanced by Artificial Intelligence, forming the Artificial Intelligence of Things (AIoT) [1]. This integration enables data collection, real-time decision-making, and autonomous action, offering potential benefits across industry, society, and remote environments. However, AIoT introduces challenges of scalability, adaptability, resilience, and sustainability that require robust and ethical strategies. Traditional definitions of intelligence emphasize acquiring and applying knowledge for problem-solving and adaptation [2,3], yet they often fail in uncertain and dynamic environments. Therefore, intelligence is here reframed as the capacity to “efficaciously enhance future possibilities under uncertainty by identifying patterns and learning from them,” prioritizing flexibility over fixed knowledge [4]. This perspective aligns with agile methodologies widely used in engineering applications. Collective intelligence is commonly defined as shared intelligence emerging through collaboration [5], but in complex systems it must also manage uncertainty, evolving objectives, and interdependent components. We therefore define collective intelligence as a dynamic network of processes enabling autonomous agents to recognize relevant information, filter and organize it, develop strategies, and continuously adapt. This aligns with Yolles and Fink’s concept of process intelligence [6] and Simon’s system hierarchy model [7], where intelligence emerges across multiple nested levels of organization. Empirical evidence for such multi-scale organization appears in recent analyses of Large Language Models (LLMs) showing atomic, cerebral, and galactic structural scales [8].

To model this form of collective intelligence, we adopt the metacybernetic paradigm [9], extended through Mindset Agency Theory (MAT) [6,9,10]. In this framework, emergent collective intelligence is formed by recursive interactions among autonomous agents. MAT builds on Varela’s autopoiesis [11,12], Schwarz’s extensions [13], and Piaget’s learning model [14], describing how process intelligences self-organize across hierarchical control structures. In metacybernetics, living systems and artificial ones, are understood generically rather than strictly biologically [15], and symbolic systems such as Artificial General Intelligence (AGI) may express life-like behaviors through autonomous information processing [16]. Core cybernetic principles—reflexivity, self-regulation, and adaptability—are already visible in AIoT applications. Digital twins provide real-time reflexive feedback loops [17,18]; smart grids demonstrate homeostatic balance [19,20]; and precision agriculture illustrates adaptive learning over time [21]. Higher-order cybernetics further supports learning and co-evolution with human users, as shown in adaptive tutoring systems [22] and sensory augmentation projects such as Cyborg Nest’s North Sense implant. These developments mirror synthetic living systems. Swarm robotics demonstrates emergent coordination like social insects, while adaptive architectural systems regulate conditions automatically, reflecting ecological efficiency principles. However, ethical considerations, especially privacy and autonomy, remain essential when such systems affect human lives. Despite progress, current IoT systems still struggle with resilience under uncertainty in multi-agent environments. To address this gap, this paper introduces COgITOR [23,24], a novel fourth-order cybernetic physical system modeling cyberfluid media, fluidic systems containing agents with cybernetic properties. COgITOR incorporates reflexive learning, holonomic memory [25], and energy harvesting [26]. While originally inspired by Pavlovian conditioning [25], it now aligns more closely with Piagetian Operative Intelligence and autopoiesis [11,27]. Building on this foundation, COgITOR is extended into Cogitor5, a fifth-order cybernetic model in which collective agency emerges from dynamic adaptive processes. Cyberfluid media modeled through Cogitor5 offer a new approach for designing scalable, autonomous, resilient AIoT systems capable of continuous adaptation.

The remainder of this paper presents: the principles of COgITOR; the (meta)cybernetic mechanisms supporting Cogitor5; the lower order cybernetic systems; the Cogitor5 model; the agency characterization through MAT. Finally, we compare Cogitor5 to a swarm intelligence model in terms of adaptability to illustrate how future IoT systems may evolve and reorganize under changing conditions.

2. COgITOR

COgITOR is an advanced cybernetic framework designed as a fourth-order regenerative agency, capable of dynamically responding to a diverse range of stimuli, particularly of electric nature, and includes conventional liquids as solvents, solid nanoparticles as dispersoids, and eventually gels, all confined within a soft artificial skin [28]. Its adaptability allows COgITOR to function effectively in complex and unpredictable environments. Based on the substrate principle of quantization, COgITOR emulates quantum-inspired behaviors by representing a complex adaptive system populated by nanoparticle agents interacting within a fluid medium. These agents remain autonomous and distinct, contributing to agency functionality by forming a responsive field of interaction that adapts to environmental demands. Furthermore, their nanometric dimensions induce the quantization of their electronic wavefunctions and opens the door to quantum phenomena [29], particularly when using magnetic nanoparticles whose spins are free to fluctuate at room temperature [30]. Such random arrangement of spins can enable average field anisotropies and can convey collective behavior, one example of which is the collective response of the magnetic colloid to an external magnetic field, well known to occur in ferrofluids [31].

Inspired by biological cells, COgITOR mimics the decentralized, self-regenerative capacities observed in cellular systems. Each agent operates independently while engaging in sophisticated inter-agent communication, akin to intercellular interactions, enabling the system to dynamically self-organize. Operating at the nanoscale, agents’ behaviors are influenced, as said, by quantum effects. This unique dynamic enables agents to perform a range of tasks, including pressure sensing [32], computation [33], data storage [34], and energy harvesting [35]. When integrated with specialized electronics, COgITOR’s infrastructure supports advanced robotics and a wide array of applications that rely on precise and adaptive functionality.

Changes at the nanoscale ripple through the COgITOR system, enhancing adaptability through non-linear dynamics. Reflexive feedback loops further bolster this adaptability, allowing COgITOR to self-repair and maintain efficacy even in extreme conditions, such as intense magnetic fields and ionizing radiation, conditions found in space [36]. Its self-healing architecture ensures resilience by fostering coevolutionary dynamics among agents, creating an ecosystem-like stability characterized by emergent properties like those observed in natural ecological systems.

COgITOR is designed for volatile environments through a distributed, self-healing architecture that enhances fault tolerance. Its agents interact across multiple scales, where nanoscale changes can trigger significant system-level responses. Quantum phase coherence enables the information carried by these agents to self-organize into stable configurations resembling ecological resilience, with transitions guided by discrete quantum energy states. Small variations in nanoparticle energy can alter magnetism, conductivity, or optical behavior, influencing agency-level dynamics [36]. This cross-scale sensitivity reflects the deterministic chaotic and fractal structure where localized perturbations produce wide-ranging adaptive effects. A key emergent feature of these quantum-driven interactions is coevolution. Different colloidal and nanoparticle agent species adapt together, forming a neo-ecosystem rather than a simple environment. Like natural ecosystems, this mutual shaping fosters stability and resilience, allowing the system to function under uncertainty. Because the system operates through nonlinear dynamics, small disturbances may produce major effects, making traditional prediction difficult. Reflexivity becomes essential: COgITOR can amplify or suppress changes to maintain adaptive balance. Through these feedback processes, the system reinforces its own resilience. These behaviors reflect neo-ecological dynamics incorporating entanglement, superposition, and mutual quantum-state modification. The result is emergent behavior, properties arising from collective interactions rather than individual agents. This strengthens system stability and supports rapid adaptation, while also enabling the development of new functional properties that improve long-term viability under changing conditions.

3. Metacybernetics

Cybernetics offers a suitable framework for modeling relationships within complex systems composed of diverse interacting entities [37]. Neoecology builds on third-order cybernetics [13,38] to examine how multiple agencies, each formed by populations of related agents, interact with one another and their surrounding environment. Rather than viewing ecosystems purely biologically, neoecology treats them as complex adaptive systems with evolving behaviors. Within this framework, a population of structurally or functionally similar agents constitutes an agency population, and distinct characteristics allow sub-populations to be understood as species. Neoecosystems extend this idea by including both biological and artificial adaptive entities, reflecting greater system complexity.

Agencies may exist within larger neoecosystems or form neoecosystems themselves. They operate autopoietically, following Maturana and Varela’s definition of self-production [27], meaning each agency maintains its identity while adapting to changing conditions. Studying a single agency reveals internal adaptive mechanisms and how these interact with external stimuli, illustrating continuous evolution rather than static structure. Cybernetic analysis examines how information flows shape control, behavior, and trajectories. Neocybernetics extends this by focusing on emergent behaviors and feedback loops in complex adaptive systems, emphasizing self-organization. Metacybernetics further develops the theory by distinguishing cybernetic orders and linking them to different ontological levels. Ontology here is divided into ordinary and fundamental domains [39]. Ordinary ontology concerns tangible, physical entities, while fundamental ontology concerns abstract structures, including meaning and consciousness. This distinction supports broader reasoning about both material systems and the principles underlying them.

Bhaskar’s critical realism [40] provides a way to relate these ontological layers through concepts such as transcendental realism, which differentiates the observable from the underlying mechanisms that generate it. In this model, reality is stratified: the superstructure corresponds to observable phenomena grounded in ordinary ontology, while the substructure represents the deeper causal mechanisms aligned with fundamental ontology. Together, these levels form a hierarchical system in which substructures govern possibilities and constraints, and superstructures express activity within those boundaries. Critical realism, therefore, enables analysis of empirical reality (what can be measured), actual reality (what exists independent of observation), and the hidden causal mechanisms that connect them. This layered perspective is essential for understanding neoecological and cybernetic systems, where both tangible and intangible structures shape behavior.

4. Cybernetic Orders and Metacybernetic Model

The word cybernetics has a proto-Indo-European origin, and recalls the concept of governing; its first documented occurrence goes back to Plato (Πλάτων, 346 B.C.), in ancient Greek: κυβϵρνητική τεχνη, literally the art of the pilot. Today it is understood as an interdisciplinary field focused on control, communication, and reflexivity in information flow [41]. In complex systems, cybernetics helps describe adaptive, non-deterministic interactions and their evolution. Different orders of cybernetics provide layered perspectives, with parsimony guiding which level best explains a system’s behavior.

First-order cybernetics, introduced by Wiener [41], focuses on agency interacting with external control mechanisms. The referent system R(1) operates in a closed loop without self-awareness, responding to environmental parameters such as temperature or humidity (see Figure 1). Behavior emerges at the superstructure, while rules and constraints exist in a deeper substructure. System interaction occurs through autopraxis: in posterior autopraxis the agency responds outward, modifying the environment; in anterior autopraxis, the environment responds inward, modifying the agency. A river basin provides a clear example: topography, geology, and hydrology regulate water flow and erosion patterns.

Second-order cybernetics, introduced by von Foerster [42], shifts focus to the role of the symbolic observer. Alongside the operative system R(1) exists a dispositional system R(2), coupled through Ѳ(1). This structure allows reflexive adaptation but not self-consciousness. Artificial Intelligence (AI) is a typical second-order example.

Third-order cybernetics, introduced by Lepskiy [38], incorporates sustentation, emphasizing homeostatic maintenance and viability through recursive adaptive learning. The sustentative system R(3) adds capacity for long-term development, co-evolution, and process intelligence. Applications include modeling biological stem cells, market systems, organizational behavior, viruses, and potentially Artificial General Intelligence (AGI).

Fourth-order cybernetics concerns metanoesis, a transformational shift in system identity and operational logic, introduced into physical systems by Chiolerio [28]. It regulates the Ѳ(2) system, enabling automorphosis, a structural self-reformulation (see Figure 2). A candidate fourth-order example is the COgITOR colloidal system.

The abstract cybernetic subsystems can be directly mapped onto common AIoT architecture patterns: sensing corresponds to anterior processes, actuation corresponds to posterior processes, and the four cybernetic system layers align, respectively, with execution, configuration, maintenance, and learning functions found in modern IoT systems.

When COgITOR is placed within a fifth-order metacybernetic framework, the resulting model, Cogitor5, extends cybernetic analysis to systems capable of recursive transformation, reflexive awareness of adaptation, and evolution across multiple ontological layers.

5. A Fifth-Order Cybernetic Agency: Cogitor5

In this framework, the focus is on the fifth-order cybernetic agency. Each level is represented as a referent system R(n), and at the highest level, R(5) satisfies the Holographic Principle [43]. In this model, the highest system acts as a holographic boundary containing only the essential information needed to regulate and coordinate lower systems, an application of the Principle of Information Parsimony. This prevents overload, supports efficient communication, and enables coherence across the hierarchy. Each level contains specialized agents responsible for local adaptive control. Fifth-order agencies operate as intangible information fields that regulate synchronization and coherence. Drawing on von Neumann, Frieden, and informational realism [44], Cogitor5 uses information as a fundamental construct, enabling interaction across classical and quantum domains. Using the Fisher Information Field Theory, the system’s phase space is treated as the holographic interface where quantum and classical behaviors converge.

Under the Principle of Information Parsimony, information flow can be formalized as a mapping $f:\mathsf{\Omega }\to {\mathbb{R}}^k$, where $\mathsf{\Omega }$ is the high-dimensional joint state space of all agents and environmental variables, and ${\mathbb{R}}^k$ is a reduced space of essential control variables required to maintain system coherence. Only the minimal information necessary for viable control is retained.

The highest-order layer is called concordance, a conceptual meta-field enabling control (see Figure 3). Analogous to bosonic coherence in lasers [45,46], concordance maintains phase alignment across the system, simulating entanglement, superposition, and collective behavior. When operating toward classical conditions, R(5) reduces to a parametric environment governing autosynesis; it introduces new behavioral possibilities beyond the classical superstructure. Within this framework, State flexibility is maintained: agents may remain in multiple potential states until conditions select one. Process intelligences support these dynamics. Autosynesis enables self-integration of information across levels; autopraxis, autopoiesis, autogenesis, and automorphosis enable autonomous action, self-maintenance, evolution, and deep structural reformulation. These processes operate with both posterior orientation (responding to input) and anterior orientation (anticipating future states). A useful analogy is Schrödinger’s cat: system agents may exist in multiple potential states until contextual conditions actualize one. Through emergence and symmetry-breaking [47,48], decisions form not from local rules but from dynamic interaction across levels (global). Thus, the concordance system of Cogitor5 functions as a hybrid interface bridging quantum-inspired and classical behavior (lower order systems). By combining holographic organization, informational parsimony, and quantum-inspired coherence, it supports adaptive learning, synchronization, and system-wide resilience. This approach offers a conceptual foundation for future AIoT systems capable of integrating quantum devices and operating in coherent, adaptive states.

The concordance system R(5) functions as a hybrid agency layer combining quantum-inspired and classical behavior. Drawing on von Neumann [49] and Frieden [50], it treats information as a fundamental entity and supports coherence and synchronization across lower cybernetic levels. Its structure balances order and chaos, enabling both stability and innovation. The concordance layer optimizes a Fisher-information–based functional defined over the global control parameters, selecting the configuration that maximizes information efficiency under viability constraints. In Figure 3, this optimization corresponds to the decision node at R(5), where the information field determines the outgoing synchronization signal.

A key mechanism is autosynesis, the self-integration of system information. Through ongoing feedback loops, autosynesis maintains coherence across levels, supports autonomous information processing, and enables adaptive learning. This capability is especially relevant for future AIoT systems that may embed quantum devices and operate in coherent states. Autosynesis acts as a cross-level integrator rather than a competing process. Anterior and posterior expressions of autopraxis, autopoiesis, autogenesis, and automorphosis produce candidate updates based on either predictive or reactive conditions. Autosynesis then assimilates these contributions and generates a coherent representation of system state, which is subsequently used by the concordance system to determine the next global decision.

The behavior of the concordance system recalls quantum field theory, where bosons share the same quantum state and act collectively. This analogy helps explain how system agents synchronize to improve collective decision-making. Concepts such as superposition, illustrated by the paradox of Schrödinger’s Cat, describe how agents may exist in multiple potential states before conditions select one. Emergence plays a central role by dynamically reshaping constraints and possibilities as interactions unfold. Simulated entanglement ensures that decisions at higher levels influence behavior across the system, while environmental inputs are assimilated at lower levels to refine adaptation. As a hybrid interface, the concordance system blends quantum-like coherence with classical operational functionality. Embedded meta-strategies evaluate multiple pathways, enabling flexible responses under uncertainty and outperforming traditional deterministic approaches.

Multiple process intelligences operate within this model, in two modes: posterior orientation, responding to incoming data; anterior orientation, anticipating future scenarios through reflexive feedback. Together, these mechanisms enable the system to refine its strategies continually, enhance coherence among agents, and maintain adaptability under changing conditions. The result is a resilient cybernetic architecture capable of navigating complexity and supporting emergent decision-making.

6. Agency Character

Agency character can be understood through three functional dimensions identified by Yolles [10]: cognition, affect, and conation. Cognition involves reasoning, information processing, and decision-making, enabling an agent to interpret its environment, learn, and adapt. Affect refers to sentience, including subjective responses, emotional valence, and arousal, shaping how the agent reacts to internal and external stimuli. Conation functions as the integrating dimension, balancing cognition and affect to maintain coherence, stability, and long-term adaptability. Within this framework, conation provides direction, while affect supplies momentum, influencing how cognition operates during decision-making. Their integration supports purposeful rather than purely reactive behavior, allowing agents to remain adaptive in changing environments. Applied to human systems, this model suggests consciousness may emerge from the interaction of cognition and affect, as discussed by Yolles [51]. Affect supplies emotional context, while cognition interprets information, together creating awareness and meaningful engagement with the environment. Through recursive interactions of cognition, affect, and conation, agency develops adaptive consciousness capable of decision-making under uncertainty and long-term behavioral refinement. This supports stability and evolution in complex environments. Consciousness, however, is not an absolute state; as noted by Bitbol and Luisi [52] and Bielecki [53], it exists across levels—ranging from minimal forms to forms approaching collective consciousness (see

Table 1. Levels of consciousness.

Level of Consciousness	Hierarchy [52]	Hierarchy [53]	Relationship
1	Null pre-conscious. Devoid of internalization. Lack of awareness, but capable of cognition.	Reflexive. A generic living system can only create behaviors that directly support existence and remove threats	Null preconscious occurs prior to reflexive states since, in the former, threats cannot be recognized.
2	Limited consciousness. Integration of environmental factors.	Limited consciousness. Integration of environmental factors. Associative. Able to undertake simple analysis of direct cause-and-effect relationships.	Limited consciousness occurs at a stage before the associative stage, the former being devoid of analytic ability.
3	Enduring modifications in self-production. Stable dynamic support provided able to deliver strongly anticipative behavior.	Conscious. Can model complex cause-and-effect chains, with a conditional option permitting future events variants and an ability for complex strategies of activity.	Enduring modifications in self-production is approximated by the consciousness stage since cause–effect chains deliver a strategy that implies anticipatory adaptation.
4	More complex changes that influence behavior. Involves observation of the exterior, but without awareness of an external independent world.	Self-consciousness. Epistemic perspective can change, with awareness of the existence of conscious goals perhaps devoid of proven reliable criteria.	More complex changes are prior to self-consciousness since the proof requires awareness and access to the outside independent world.
5	Collective social consciousness	The hypothetical omniscient stage, with proven criteria and proof of the reliability to use it.	Collective consciousness is likely equivalent to omniscient if one considers that proof is a social phenomenon.

).

The relationships between regenerative biological systems and Cogitor5 can be explored conceptually, as illustrated in

Table 2. Comparing Stem Cell and Cogitor5 Distinctive Actions in Aspects and the Resulting Consciousness, and IoT.

Aspect	Description	Stem Cell Agency	Cogitor5 Agency
Cognition	Sapience functionality: Reasoning, information processing, and decision-making regarding internal and external environments.	Processes molecular signals to inform trajectory decisions such as division and differentiation. These processes are shaped by both internal and external signals, embodying a basic form of biological decision-making influenced by environmental and cellular data. Cells display adaptive responses that guide differentiation and proliferation through intrinsic biological programs, supported by agents like signaling molecules, nucleic acids, and structural proteins, and cellular structures like organoids.	Processes environmental data (e.g., energy, chemicals, physical stimuli) through collective interactions. They exhibit emergent cognition-like behavior by dynamically reconfiguring and optimizing responses to environmental demands. Reflects a basic form of real-time data processing and decision-making, enabled by self-organization through inherent algorithms derived from agent interactions.
Affect	Sentience functionality: Awareness and subjective experiences affecting responses to the environment through valence and arousal.	Reacts to environmental stimuli, such as tissue damage, with responses that reflect primitive emotional regulation. For example, a negative stimulus like tissue damage triggers division and repair (a positive adaptation). The intensity of response (arousal) is influenced by damage severity and the surrounding microenvironment. Biochemical signals guide whether stem cells differentiate or proliferate, indicating their readiness to respond to stimuli.	Dynamically responds to environmental changes (e.g., energy inputs, chemical cues), causing organizational change. Valence indicates agency stability or instability, while arousal may correspond to the intensity of adjustments. Self-regulation may be based on external stimuli, reflecting a primitive awareness of their environment. Adaptation responds to the intrinsic properties of stimuli, influencing immediate responses with no long-term behavioral modification capabilities.
Conation	Balancing impulse between cognition and affect: drives the system toward maintained viability rather than mere uniformity.	Maintains viability by balancing biochemical and physiological processes, ensuring effective tissue repair and operational continuity. This balance sustains long-term adaptation to environmental needs without enforcing uniformity.	Implements purposeful coherence through distributed state synchronization. Agents update local variables relative to shared references (imbalance flags, global mode indicators, reference trajectories), enabling resilience and continuity in dynamic environments such as industrial IoT or smart grids.
Consciousness	Emergent Property: Interaction between cognition and affect mediated by conation, involving awareness and complex responses to the environment.	Exhibits a primitive awareness, demonstrating responsiveness to environmental signals contributing to survival and adaptation, albeit lacking higher-order consciousness. Maintaining long-term adaptability through dynamic differentiation, supported by biological agents like biomolecules and organoids, ensures cohesion between differentiation processes and environmental needs.	Reflects emergent “cyber-consciousness,” from dynamic interaction between the aspects. Enables real-time cognition (data processing) and adaptive learning, facilitating complex adaptive responses to environmental changes. Adaptive learning enables self-renewal and sustainability. Shows spontaneous organization and interconnectedness.

. This comparison examines how two distinct forms of agency: Stem Cell Agency and Cogitor5 Agency demonstrate adaptation, responsiveness, and coherence under complexity. Stem Cell Agency represents a biological model where decision-making, environmental responsiveness, and self-regulation occur through biochemical pathways. Stem cells exhibit primitive cognition, affect-like responsiveness, and coherence by adjusting behavior to organismal needs. Although not conscious in a human sense, they demonstrate awareness through adaptive responses to injury and signaling processes. In contrast, Cogitor5 Agency represents a cybernetic form of agency operating through networked interactions. Instead of biochemical mechanisms, its behavior emerges from real-time information exchange among agents. It exhibits self-organization, adaptive restructuring, and long-term stability. The resulting emergent behavior may be interpreted as cyber-consciousness, arising from collective processes rather than a single component.

The comparison suggests that adaptive agency, self-regulation, and emergent awareness are not exclusive to biological systems. Rather, they may appear in synthetic systems when supported by cognition, affect-like response dynamics, and coherence.

This framework extends to the Internet of Things (IoT) by aligning its core functionalities with the three agency aspects defined in Mindset Agency Theory [10,54]:

• Cognition: IoT systems collect and process environmental data through sensors, networks, and algorithms. Real-time responses (like smart thermostats adjusting temperature or machine-to-machine communication enabling coordinated behavior) reflect operational cognition.
• Affect: although non-sentient, IoT systems exhibit affect-like responses by adjusting behavior according to valence (stability) and arousal (response intensity). Examples include smart irrigation systems modifying water flow based on soil moisture or adaptive traffic systems responding to congestion.
• Conation: this acts as the balancing impulse between cognition and affect, driving the system toward maintained viability. In IoT this appears as frameworks that sustain operational continuity: industrial IoT balancing production and maintenance cycles, or smart grids dynamically reallocating energy to preserve resilience. Key state variables are synchronized across the system through a broadcast mechanism implemented at the concordance layer. The distributed state is compressed into a low-dimensional signal—imbalance flags, global mode indicators or reference trajectories—which is redistributed to all agents. Each agent updates its local variables relative to this shared reference, enabling purposeful coherence without enforcing identical states.

Together, the comparison across stem cells, Cogitor5, and IoT highlights a broader definition of agency applicable to biological and synthetic systems. It shows how cognition, affect, and conation enable adaptive function, emergent behavior, and coherence across diverse forms of complex organization.

Table 3. Comparative Control States for Stem Cells and Cogitor5 Agency.

System	Stem Cell Agency	Cogitor5 Agency
Operative (1st order)	Executes biological tasks such as cell division and differentiation. Maintains structured action sequences based on genetic information, biochemical processes, and interactions among cellular agents (e.g., proteins, nucleic acids) and organelles.	Exhibits self-organizing behavior through the interaction of species of colloidal agents and nanoparticles. These interactions form organized structures and pathways responding to stimuli, leading to emergent actions. Autopoiesis provides the foundation for self-organization and adaptive learning, allowing real-time adjustments based on environmental changes.
Dispositional (2nd order)	Regulates cell behavior and development through autopoiesis. Stores structured object relationships (e.g., genetic instructions, molecular pathways) that ensure adaptive growth and specialization in response to the environment.	Based on adaptive routines enabled by built-in regulatory frameworks or external stimuli manifested as information-based structures. Functionalized colloidal agents and nanoparticles act as dynamic sensors or catalysts, allowing regulatory behaviors to emerge from the system’s interactions. Adaptive learning algorithms further refine these responses, continuously promoting coherence and adaptability.
Sustentative (3rd order)	Regulates homeostasis through structured knowledge of internal biochemical states (e.g., biochemical signaling pathways, feedback loops involving organelles such as mitochondria and endoplasmic reticulum). Adjusts via autogenesis to maintain equilibrium.	Stability arises from the physical and chemical equilibria established by species of colloidal agents and nanoparticles. While they can adjust to external forces, they require engineered adaptive learning mechanisms to mimic the inherent adaptability of biological systems. This creates a feedback loop that enhances stability and adaptability.
Metanoetic (4th order)	Regulates higher-order transformation and reconfiguration through automorphosis, where organelles can reorganize and cells can evolve or differentiate into more complex forms driven by internal signals and interactions among biological agents.	Transformation in Cogitor5 agency is informed by external stimuli and adaptive routines that enable reconfiguration of the system’s internal architecture. Incorporating feedback mechanisms and engineered adaptive learning, these systems reflect fourth-order principles, allowing for continuous evolution and reorganization in response to environmental dynamics.
Concordance (5th order)	Theorized as Integrating quantized interactions at the stem cell level, allowing decision-making processes influenced by discrete quantum signals. It has been postulated to use quantum coherence to coordinate developmental outcomes, reflecting a higher order of complexity in stem cell behavior [55].	Coordinates interactions among colloidal agents and nanoparticles through quantized signals, utilizing quantum coherence to synchronize agent interactions and ensure adaptive behaviors emerge in response to both internal configurations and external influences. The system leverages quantum principles and adaptive learning mechanisms to enhance responsiveness and continuous evolution.

compares Stem Cell Agency and Cogitor5 Agency as living systems across five levels of control: operative, dispositional, sustentative, metanoetic, and concordance. These layers represent how each system organizes itself, adapts to its environment, and maintains coherence through different forms of process intelligence. The framework can be extended to include the Internet of Things (IoT) by aligning IoT behavior with the biological characteristics of stem cells and the cybernetic properties of Cogitor5. At lower levels, IoT systems parallel stem cells through task execution, feedback-driven adaptation, and real-time interaction—functions central to both IoT networks and biological cell behavior. As complexity increases, IoT systems can be viewed analogously to higher-order cybernetic structures, capable of self-organization, coordinated interaction, and emergent behavior. At the highest levels, speculative alignment with concepts such as quantum coherence suggests a future in which IoT systems may achieve integration and synchronization comparable to the mechanisms hypothesized in Cogitor5. This progression highlights how IoT, like biological and cybernetic systems, may evolve from basic operational capability to higher-order adaptive intelligence and coherence.

Table 4. Comparing Stem Cell and Cogitor5 Agency and IoT in their Process Intelligence Actions.

Process Intelligence	Stem Cell Agency	Cogitor5 Agency	Internet of Things (IoT)
Autopraxis	Autonomous acquisition of environmental data through receptors that sense signals and chemical cues, important for differentiation and development. The extracellular matrix (ECM) provides context and support for these processes [56].	Mechanisms could involve sensors within Cogitor5 agents that gather environmental data and relay it to decision-making. This manifold informs trajectory formation and enhances adaptability among colloidal and nanoparticle agents. An extraCogitor5 matrix could support these interactions.	IoT devices, equipped with sensors, autonomously gather data about environmental conditions (e.g., temperature, humidity, light). Data networks act as the “matrix,” providing a context for devices to interpret inputs and adjust their operations. Examples include smart home systems and environmental monitoring networks that autonomously adapt based on real-time data.
Autopoiesis	Delivers mechanisms like cell proliferation and differentiation, driven by intrinsic signals, enabling self-maintenance and regeneration [57,58]. Scaffold Integrity aids in maintaining cellular structures [59].	Agents engage in self-generative behaviors that promote stability and resilience by autonomously forming complex structures in response to their environment. These interactions enable dynamic responses to external changes, mirroring biological self-organization processes that lead to functional diversity [60].	IoT systems exhibit self-organizing properties, such as networked devices reconfiguring themselves to maintain functionality during outages or disruptions. For instance, in smart grids, sensors and controllers dynamically balance power loads to ensure system stability and self-sustainability.
Autogenesis	Exhibit self-creation through processes like stemness maintenance and lineage commitment, leading to diverse cell types that adapt to environmental needs [61].	Processes could involve agents evolving into more complex structures (as collective subagencies) or functions, adapting to dynamic environments and creating novel behaviors [62].	IoT systems evolve through adaptive firmware and software updates, enabling devices to acquire new capabilities or functionalities. Examples include autonomous vehicles integrating machine learning updates to improve navigation and safety features.
Automorphosis	By processes of differentiation, leads to functional specialization and enhancement of adaptability [63].	Might manifest as the reconfiguration of agent interactions, allowing them to adapt their behavior and functionality in response to varying conditions [64].	IoT systems exhibit functional specialization through dynamic configuration, such as edge computing nodes reassigning tasks to optimize performance based on real-time conditions. For instance, IoT-enabled manufacturing lines reconfigure robotic operations to respond to changing production requirements.
Autosynesis	Integration of intracellular signaling pathways (mechanical, genetic, biochemical) to create unified cellular decisions. Represents sense-making across molecular subsystems.	Provides system-wide integration of information across referent layers via coherence functions. Acts as the mechanism harmonizing posterior (reactive) and anterior (predictive) process intelligences.	Unifies analytics, device state, and orchestration layers into a coherent decision fabric. Example: cloud–edge–agent alignment enabling system-wide optimization (e.g., digital twins ensuring consensus states across distributed nodes).

summarizes four forms of process intelligence (autopraxis, autopoiesis, autogenesis, automorphosis, and autosynesis) that support autonomous function, adaptation, and ongoing development in both stem cells and Cogitor5 Agency systems. This framework can also be applied to the IoT. Autosynesis is reflected in IoT sensing and data acquisition, which parallels stem cell receptors and the role of the extracellular matrix—both providing contextual environmental awareness. Autopoiesis aligns with the self-maintaining and resilient nature of IoT networks, which reorganize and recover much like biological systems. Autogenesis appears in IoT system evolution through software updates, reconfiguration, and machine learning, mirroring adaptive development seen in stem cells and Cogitor5 agents. Automorphosis corresponds to IoT’s capacity for functional adaptation and restructuring, similar to biological differentiation under environmental influence. Together, these parallels show how IoT systems can emulate (and extend) the adaptive behavior of biological and synthetic agents across increasing levels of complexity.

7. Faulting Grid Case Study: IoT-Applied Scenario

In order to provide a practical comparison of the Cogitor5 model applied to the IoT domain, we have performed a Matlab^® (https://ww2.mathworks.cn/products/matlab.html) simulation to assess whether our framework with its five hierarchical orders and process intelligences outperforms a conventional swarm-intelligence baseline or not, in metrics such as convergence speed, resilience to faults [65,66,67], and energy efficiency. The simulation contains a network of N nodes connected to the electric grid, that could represent renewable energy sources, storage units, and loads. The task is to dynamically balance power supply and demand in real time under fluctuating renewable generation, with sudden load spikes or node failures, which is a frequent scenario happening in the renewable energy landscape. The objective function is to minimize the total supply–demand imbalance and stabilize the network in the shortest time, while maximizing the energy available. To do so we have compared a baseline swarm intelligence model, in other words a standard IoT network, where the architecture is flat, in which homogeneous agents use local consensus algorithms [68,69]. This system belongs to the first-order cybernetics, featuring single-layer reflexive feedback. Each agent in the baseline model senses the local supply/demand, communicates with neighbors to estimate the global state of the network, adjusts the local output via proportional control and swarm heuristic. The higher order model, based on the Cogitor5 scheme, features a hierarchical layer control, each featuring specialized agents and process intelligences. The operative layer (1) executes local sensing and actuation (autopraxis), the dispositional layer (2) maintains a short-term memory of recent states and regulates each agent behavior (autopoiesis), the sustentative layer (3) monitors homeostatic bounds (e.g., voltage/frequency limits) and triggers the self-creation of new operational modes switching from grid-tied to islanded mode (autogenesis), the metanoetic layer (4) reconfigures the network topology or agent roles in response to major faults via reassigning storage nodes to compensate for generator failure (automorphosis), and the concordance layer (5) uses Fisher Information or phase-space parsimony to broadcast global synchronization signals ensuring coherent convergence via self-integration of distributed states (autosynesis). When a decision is generated at the concordance layer R(5), it is propagated downward through the system hierarchy as a compact set of updated global parameters or mode directives. Each lower layer integrates these parameters according to its function: the metanoetic layer translates them into structural reconfiguration instructions, the sustentative layer updates homeostatic bounds, the dispositional layer adjusts local rules and controller gains, and the operative layer finally applies them at the level of individual agent actuation. This top-down cascade ensures that higher-order decisions become coherent behavioral adjustments across the multi-agent system.

If multiple processes propose conflicting updates during the same execution tick, resolution follows a priority hierarchy: viability and safety constraints dominate performance preferences, metanoetic reconfiguration overrides dispositional adjustments, and the concordance system performs the final aggregation. The resulting unified update is then applied system-wide.

The higher complexity of the Cogitor5 model is paid back by the higher efficiency in a real case scenario as shown in Figure 4, where imbalance measures how far the network is from satisfying supply equals demand at a given time (lower is better), and the total energy is a proxy for how much aggregate actuation/output the network is producing. The computation is performed in such a way: at each time step, the mismatch per node is computed as supply − demand (an N × 1 vector). The scalar imbalance is then the Euclidean norm of that vector: ‖supply − demand‖₂. This aggregates all local mismatches into a single global error metric. It goes to zero when every node’s local balance is met or when network-wide supply equals demand. It represents a dynamic performance indicator of how well the IoT model has coordinated to meet the load-balancing objective. Faster decay of imbalance after node failures means better convergence and resilience. The total energy is ∑ᵢ |supplyᵢ|. Overall the simulation shows that Cogitor5 features a 30% faster initial stabilization, a 2–5% lower imbalance, and a higher 2–5% efficiency.

Modern IoT wireless technologies such as Chirp Spread Spectrum (CSS) [70,71] and its variants, including the recently proposed Dual-Mode Time Domain Multiplexed Chirp Spread Spectrum (DM-TDM-CSS) [72,73], offer an engineered substrate for device interconnection. CSS/DM-TDM-CSS excel in robust physical-layer communication, providing long-range, low-power connectivity for distributed sensor/actuator networks. The DM-TDM-CSS variant further improves spectral and energy efficiency preserving low cost and low complexity. In contrast, the Cogitor5/AIoT-cybernetic architecture emphasizes multi-scale adaptive organization, information-parsimony, reflexivity, emergent agency and hierarchical control, features that go well beyond simple data transmission. Here the agents are conceived as self-organizing, learning and evolving entities capable of collective intelligence, adaptive reconfiguration, and context-aware decision-making across multiple orders of control.

8. Conclusions

Cybernetic theory has progressed from basic system–environment interaction in first-order cybernetics to self-regulating and adaptive structures in second- and third-order frameworks. Fourth-order cybernetics, introducing concepts such as metanoesis and automorphosis, focuses on systems capable of deep internal transformation. Fifth-order cybernetic agencies, such as Cogitor5, extend this evolution by integrating quantum-inspired and classical principles, enabling synchronized, adaptive responses to complex environments. This progression supports the development of advanced AIoT systems capable of real-time evolution and precision adaptation. Within this framework, agency is defined by three dimensions: cognition, affect, and conation. Cognition describes reasoning and information processing; affect reflects responses and regulation influenced by emotional-like states; and conation integrates the two, maintaining coherence, stability, and adaptability. Their interaction enables functional agency capable of responding to both internal dynamics and external conditions. Consciousness arises from the synergy between cognition and affect, enabling meaningful engagement with the environment. Recursion among cognition, affect, and conation allows adaptive consciousness to evolve over time, supporting resilience and long-term viability. A comparison between Stem Cell Agency and Cogitor5 Agency shows that agency can appear in both biological and cybernetic systems. Stem cells express primitive agency through biochemical regulation, while Cogitor5 operates through real-time computation and self-organization. This implies that agency is not confined to living organisms but can emerge wherever cognitive, affective, and integrative dynamics are present. IoT systems also reflect these dimensions. Their data processing and real-time decision-making demonstrate cognition, their adaptive responses to environmental conditions resemble affect, and their capacity to maintain systemic coherence imitates quantum phenomena. Although IoT does not possess biological consciousness, its behavior reflects forms of adaptive agency. Viewed through metacybernetic theory, agency becomes a property of complex adaptive systems—biological or synthetic—where cognition processes information, affect modulates responsiveness, and conation provides orientation and coherence. Together, these dimensions support system viability and adaptive evolution across diverse forms of organization.