In the present study, 64 users were asked to convey eight distinct emotion to a humanoid Nao robot via touch, and were then asked to evaluate their experiences of performing that task. Large differences between emotions were revealed. Users perceived conveying of positive/pro-social emotions as significantly easier than negative emotions, with love and disgust as the two extremes. When asked whether they would act differently towards a human, compared to the robot, the users’ replies varied. A content analysis of interviews revealed a generally positive user experience (UX) while interacting with the robot, but users also found the task challenging in several ways. Three major themes with impact on the UX emerged; responsiveness, robustness, and trickiness. The results are discussed in relation to a study of human-human affective tactile interaction, with implications for human-robot interaction (HRI) and design of social and affective robotics in particular.
Social robots are expected gradually to be used by more and more people in a widerrange of settings, domestic as well as professional. As a consequence, the features and qualityrequirements on human–robot interaction will increase, comprising possibilities to communicateemotions, establishing a positive user experience, e.g., using touch. In this paper, the focus is ondepicting how humans, as the users of robots, experience tactile emotional communication with theNao Robot, as well as identifying aspects affecting the experience and touch behavior. A qualitativeinvestigation was conducted as part of a larger experiment. The major findings consist of 15 differentaspects that vary along one or more dimensions and how those influence the four dimensions ofuser experience that are present in the study, as well as the different parts of touch behavior ofconveying emotions.
Affective touch has a fundamental role in human development, social bonding, and for providing emotional support in interpersonal relationships. We present, what is to our knowledge, the first HRI study of tactile conveyance of both positive and negative emotions (affective touch) on the Nao robot, and based on an experimental set-up from a study of human–human tactile communication. In the present work, participants conveyed eight emotions to a small humanoid robot via touch. We found that female participants conveyed emotions for a longer time, using more varied interaction and touching more regions on the robot’s body, compared to male participants. Several differences between emotions were found such that emotions could be classified by the valence of the emotion conveyed, by combining touch amount and duration. Overall, these results show high agreement with those reported for human–human affective tactile communication and could also have impact on the design and placement of tactile sensors on humanoid robots.
A technique for Simultaneous Planning and Action (SPA) based on Dynamic Field Theory (DFT) is presented. The model builds on previous workon representation of sequential behavior as attractors in dynamic neural fields. Here, we demonstrate how chains of competing attractors can be used to represent dynamic plans towards a goal state. The presentwork can be seen as an addition to a growing body of work that demonstratesthe role of DFT as a bridge between low-level reactive approachesand high-level symbol processing mechanisms. The architecture is evaluatedon a set of planning problems using a simulated e-puck robot, including analysis of the system's behavior in response to noise and temporary blockages ofthe planned route. The system makes no explicit distinction betweenplanning and execution phases, allowing continuous adaptation of the planned path. The proposed architecture exploits the DFT property of stability in relation to noise and changes in the environment. The neural dynamics are also exploited such that stay-or-switch action selection emerges where blockage of a planned path occurs: stay until the transient blockage is removed versus switch to an alternative route to the goal.
Several simulation theories have been proposed as an explanation for how humans and other agents internalize an "inner world" that allows them to simulate interactions with the external real world - prospectively and retrospectively. Such internal simulation of interaction with the environment has been argued to be a key mechanism behind mentalizing and planning. In the present work, we study internal simulations in a robot acting in a simulated human environment. A model of sensory-motor interactions with the environment is generated from human demonstrations, and tested on a Robosoft Kompai robot. The model is used as a controller for the robot, reproducing the demonstrated behavior. Information from several different demonstrations is mixed, allowing the robot to produce novel paths through the environment, towards a goal specified by top-down contextual information.
The robot model is also used in a covert mode, where actions are inhibited and perceptions are generated by a forward model. As a result, the robot generates an internal simulation of the sensory-motor interactions with the environment. Similar to the overt mode, the model is able to reproduce the demonstrated behavior as internal simulations. When experiences from several demonstrations are combined with a top-down goal signal, the system produces internal simulations of novel paths through the environment. These results can be understood as the robot imagining an "inner world" generated from previous experience, allowing it to try out different possible futures without executing actions overtly.
We found that the success rate in terms of reaching the specified goal was higher during internal simulation, compared to overt action. These results are linked to a reduction in prediction errors generated during covert action. Despite the fact that the model is quite successful in terms of generating covert behavior towards specified goals, internal simulations display different temporal distributions compared to their overt counterparts. Links to human cognition and specifically mental imagery are discussed.
- A classical appraisal model of emotions extended with artificial metabolic mechanisms is presented. The new architecture is based on two existing models: WASABI and a model of Microbial Fuel Cell technology. WASABI is a top-down cognitive model which is implemented in several virtual world applications such as a museum guide. Microbial fuel cells provide energy for the robot through digesting food. The presented work is a first step towards imbuing a physical robot with emotions of human-like complexity. Classically, such integration has only been attempted in the virtual domain. The research aim is to study the embodied appraisal theory and to show the role of the body in the emotion mechanisms. Some initial tests of the architecture with humanoid NAO robot in a minimalistic scenario are presented. © Selection and peer-review under responsibility of FET11 conference organizers and published by Elsevier B.V.
In this article, a simple CPG network is shown to model early infant walking, in particular the onset of independent walking. The difference between early infant walking and early adult walking is addressed with respect to the underlying neurophysiology and evaluated according to gait attributes. Based on this, we successfully model the early infant walking gait on the NAO robot and compare its motion dynamics and performance to those of infants. Our model is able to capture the core properties of early infant walking. We identify differences in the morphologies between the robot and infant and the effect of this on their respective performance. In conclusion, early infant walking can be seen to develop as a function of the CPG network and morphological characteristics.
In this article, a generic CPG architecture is used to model infant crawling gaits and is implemented on the NAO robot platform. The CPG architecture is chosen via a systematic approach to designing CPG networks on the basis of group theory and dynamic systems theory. The NAO robot performance is compared to the iCub robot which has a different anatomical structure. Finally, the comparison of performance and NAO whole-body stability are assessed to show the adaptive property of the CPG architecture and the extent of its ability to transfer to different robot morphologies. © 2011 IEEE.
The identification of learning mechanisms for locomotion has been the subject of much research for some time but many challenges remain. Dynamic systems theory (DST) offers a novel approach to humanoid learning through environmental interaction. Reinforcement learning (RL) has offered a promising method to adaptively link the dynamic system to the environment it interacts with via a reward-based value system. In this paper, we propose a model that integrates the above perspectives and applies it to the case of a humanoid (NAO) robot learning to walk the ability of which emerges from its value-based interaction with the environment. In the model, a simplified central pattern generator (CPG) architecture inspired by neuroscientific research and DST is integrated with an actor-critic approach to RL (cpg-actor-critic). In the cpg-actor-critic architecture, least-square-temporal-difference based learning converges to the optimal solution quickly by using natural gradient learning and balancing exploration and exploitation. Futhermore, rather than using a traditional (designer-specified) reward it uses a dynamic value function as a stability indicator that adapts to the environment. The results obtained are analyzed using a novel DST-based embodied cognition approach. Learning to walk, from this perspective, is a process of integrating levels of sensorimotor activity and value.
In this article, we use a recurrent neural network including four-cell core architecture to model the walking gait and implement it with the simulated and physical NAO robot. Meanwhile, inspired by the biological CPG models, we propose a simplified CPG model which comprises motorneurons, interneurons, sensor neurons and the simplified spinal cord. Within this model, the CPGs do not directly output trajectories to the servo motors. Instead, they only work to maintain the phase relation among ipsilateral and contralateral limbs. The final output is dependent on the integration of CPG signals, outputs of interneurons, motor neurons and sensor neurons (sensory feedback).
Connectionist and bio-inspired approaches to the study of emotional learning and decision making often emphasize, or imply, an executive role for the brain whilst paying only lip service to the role of the non-neural body. In this short paper I will discuss approaches to modelling emotions that have attempted to take into account, in one form or another, the role of the body in emotional learning and decision making. More specifically, I will argue that the ‘how’ of behavioural responding and not just the ‘what’ must be factored into any learning algorithm that purports to be emotional. Furthermore, I will refer to research that has utilized abstract artificial environments designed to explore the relevance of how behaviours are carried out with a view to scaling performance to more complex, including human-based, environments.
The partial reinforcement extinction effect (PREE) is an experimentally established phenomenon: behavioural response to a given stimulus is more persistent when previously inconsistently rewarded than when consistently rewarded. This phenomenon is, however, controversial in animal/human learning theory. Contradictory findings exist regarding when the PREE occurs. One body of research has found a within-subjects PREE, while another has found a within-subjects reversed PREE (RPREE). These opposing findings constitute what is considered the most important problem of PREE for theoreticians to explain. Here, we provide a neurocomputational account of the PREE, which helps to reconcile these seemingly contradictory findings of within-subjects experimental conditions. The performance of our model demonstrates how omission expectancy, learned according to low probability reward, comes to control response choice following discontinuation of reward presentation (extinction). We find that a PREE will occur when multiple responses become controlled by omission expectation in extinction, but not when only one omission-mediated response is available. Our model exploits the affective states of reward acquisition and reward omission expectancy in order to differentially classify stimuli and differentially mediate response choice. We demonstrate that stimulus–response (retrospective) and stimulus–expectation–response (prospective) routes are required to provide a necessary and sufficient explanation of the PREE versus RPREE data and that Omission representation is key for explaining the nonlinear nature of extinction data.
Joint Action is typically described as social interaction that requires coordination among two or more co-actors in order to achieve a common goal. In this article, we put forward a hypothesis for the existence of a neural-computational mechanism of affective valuation that may be critically exploited in Joint Action. Such a mechanism would serve to facilitate coordination between co-actors permitting a reduction of required information. Our hypothesized affective mechanism provides a value function based implementation of Associative Two-Process (ATP) theory that entails the classification of external stimuli according to outcome expectancies. This approach has been used to describe animal and human action that concerns differential outcome expectancies. Until now it has not been applied to social interaction. We describe our Affective ATP model as applied to social learning consistent with an “extended common currency” perspective in the social neuroscience literature. We contrast this to an alternative mechanism that provides an example implementation of the so-called social-specific value perspective. In brief, our Social-Affective ATP mechanism builds upon established formalisms for reinforcement learning (temporal difference learning models) nuanced to accommodate expectations (consistent with ATP theory) and extended to integrate non-social and social cues for use in Joint Action.
We here present results and analysis from a study of affective tactile communication between human and humanoid robot (the NAO robot). In the present work, participants conveyed eight emotions to the NAO via touch. In this study, we sought to understand the potential for using a wearable affective (tactile) interface, or WAffI. The aims of our study were to address the following: (i) how emotions and affective states can be conveyed (encoded) to such a humanoid robot, (ii) what are the effects of dressing the NAO in the WAffI on emotion conveyance and (iii) what is the potential for decoding emotion and affective states. We found that subjects conveyed touch for longer duration and over more locations on the robot when the NAO was dressed with WAffI than when it was not. Our analysis illuminates ways by which affective valence, and separate emotions, might be decoded by a humanoid robot according to the different features of touch: intensity, duration, location, type. Finally, we discuss the types of sensors and their distribution as they may be embedded within the WAffI and that would likely benefit Human-NAO (and Human-Humanoid) interaction along the affective tactile dimension.
Robots inhabiting human environments need to act in relation to their own experience and embodiment as well as to social and emotional aspects. Robots that learn, act upon and incorporate their own experience and perception of others’ emotions into their responses make not only more productive artificial agents but also agents with whom humans can appropriately interact. This special issue seeks to address the significance of grounding of emotions in robots in relation to aspects of physical and homeostatic interaction in the world at an individual and social level. Specific questions concern: How can emotion and social interaction be grounded in the behavioral activity of the robotic system? Is a robot able to have intrinsic emotions? How can emotions, grounded in the embodiment of the robot, facilitate individually and socially adaptive behavior to the robot? This opening chapter provides an introduction to the articles that comprise this special issue and briefly discusses their relationship to grounding emotions in robots.
In this article we present a novel neural network implementation of Associative Two-Process (ATP) theory based on an Actor–Critic-like architecture. Our implementation emphasizes the affective components of differential reward magnitude and reward omission expectation and thus we model Affective-Associative Two-Process theory (Aff-ATP). ATP has been used to explain the findings of differential outcomes training (DOT) procedures, which emphasize learning differentially valuated outcomes for cueing actions previously associated with those outcomes. ATP hypothesizes the existence of a ‘prospective’ memory route through which outcome expectations can bring to bear on decision making and can even substitute for decision making based on the ‘retrospective’ inputs of standard working memory. While DOT procedures are well recognized in the animal learning literature they have not previously been computationally modelled. The model presented in this article helps clarify the role of ATP computationally through the capturing of empirical data based on DOT. Our Aff-ATP model illuminates the different roles that prospective and retrospective memory can have in decision making (combining inputs to action selection functions). In specific cases, the model’s prospective route allows for adaptive switching (correct action selection prior to learning) following changes in the stimulus–response–outcome contingencies.
In this chapter, different notions of allostasis (the process of achieving stability through change ) as they apply to adaptive behavior are presented. The authors discuss how notions of allostasis can be usefully applied to Cybernetics-based homeostatic systems. Particular emphasis is placed upon affective states - motivational and emotional - and, above all, the notion of 'predictive' regulation, as distinct from forms of 'reactive' regulation, in homeostatic systems. The authors focus here on Ashby's ultrastability concept that entails behavior change for correcting homeostatic errors (deviations from the healthy range of essential, physiological, variables). The authors consider how the ultrastability concept can be broadened to incorporate allostatic mechanisms and how they may enhance adaptive physiological and behavioral activity. Finally, this chapter references different Cybernetics frameworks that incorporate the notion of allostasis. The article then attempts to untangle how the given perspectives fit into the 'allostatic ultrastable systems' framework postulated.
Choice behaviour where outcome-contingencies vary or are prohabilistic has been the focus of many benchmark tasks of infant to adult development in the psychology literature. Dynamic field theoretic (DFT) investigations of cognitive and behavioural competencies have been used in order to identify parameters critical to infant development. In this paper we report the findings of a DFT model that is able to replicate normal functioning adult performance on the Iowa gambling task (IGT). The model offers a simple demonstration proof of the parsimonious reversal learning alternative to Damasio’s somatic marker explanation of IGT performance. Our simple model demonstrates a potentially important role for reinforcement/reward learning to generating behaviour that allows for advantageous performance. We compare our DFT modelling approach to one used on the A-not-B infant paradigm and suggest that a critical aspect of development lies in the ability to flexibly trade off perseverative versus exploratory behaviour in order to capture statistical choice-outcome contingencies. Finally, we discuss the importance of an investigation of the IGT in an embodied setting where reward prediction learning may provide critical means by which adaptive behavioural reversals can be enacted.
Emotions can be considered inextricably linked to embodied appraisals - perceptions of bodily states that inform agents of how they are faring in the world relative to their own well-being. Emotion-appraisals are thus relational phenomena the relevance of which can be learned or evolutionarily selected for given a reliable coupling between agent-internal and environmental states. An emotion-appraisal attentional disposition permits agents to produce behaviour that exploits such couplings allowing for adaptive agent performance across agent-environment interactions. This chapter discusses emotions in terms of dynamical processes whereby attentional dispositions are considered central to an understanding of behaviour. The need to reconcile a dynamical systems perspective with an approach that views emotions as attentional dispositions representative of embodied relational phenomena (embodied appraisals) is argued for. Attention and emotion are considered to be features of adaptive agent behaviour that are interdependent in their temporal, structural and organizational relations.
Research on the neural bases of emotion raises much controversy and few quantitative models exist that can help address the issues raised. Here we replicate and dissect one of those models, Armony and colleagues’neurocomputational model of fear conditioning, which is based on LeDoux’s dual-route hypothesis regarding the rat fear circuitry. The importance of the model’s modular abstraction of the neuroanatomy, its use of population coding, and in particular the interplay between thalamo-amygdala and thalamo-cortical pathways are tested. We show that a trivially minimal version of the model can produce conditioning to a reinforced stimulus without recourse to the dual pathway structure, but a modification of the original model, which nevertheless preserves the thalamo-amygdala and (reduced) thalamo-cortical pathways, enables stronger conditioning to a conditioned stimulus. Implications for neurocomputational modelling approaches are discussed.
In this chapter, we present a minimalist approach to utilizing the computational principles of affective processes and emotions for autonomous robotics applications. The focus of this paper is on the presentation of this framework in reference to preservation of agent autonomy across levels of cognitive-affective competences. This approach views autonomy in reference to (i) embodied (e.g. homeostatic), and (ii) dynamic (e.g. neural-dynamic) processes, required to render adaptive such cognitive-affective competences. We hereby focus on bridging bottom-up (standard autonomous robotics) and top-down (psychology-based dimensional theoretic) modelling approaches. Our enactive approach we characterize according to bi-directional grounding (inter-dependent bottom-up and top-down regulation). As such, from an emotions theory perspective, ‘enaction’ is best understood as an embodied and dynamic appraisal perspective. We attempt to clarify our approach with relevant case studies and comparison to other existing approaches in the modelling literature.
We present an evolutionary robotics investigation into the metabolism constrained homeostatic dynamics of a simulated robot. Unlike existing research that has focused on either energy or motivation autonomy the robot described here is considered in terms of energy-motivation autonomy. This stipulation is made according to a requirement of autonomous systems to spatiotemporally integrate environmental and physiological sensed information. In our experiment, the latter is generated by a simulated artificial metabolism (a microbial fuel cell batch) and its integration with the former is determined by an E-GasNet-active vision interface. The investigation centres on robot performance in a three-dimensional simulator on a stereotyped two-resource problem. Motivationlike states emerge according to periodic dynamics identifiable for two viable sensorimotor strategies. Robot adaptivity is found to be sensitive to experimenter-manipulated deviations from evolved metabolic constraints. Deviations detrimentally affect the viability of cognitive (anticipatory) capacities even where constraints are significantly lessened. These results support the hypothesis that grounding motivationally autonomous robots is critical to adaptivity and cognition.
Dynamical systems perspectives on emotion emphasize the importance of the regulatory interplay between brain, body and environment to adaptive behaviour. We suggest that a key facet of emotions, above all fear, consistent with this perspective lies in the allostatic regulation of constitutive/behavioural dynamics in terms of prediction and behavioural biases linking internal needs to external adaptive concerns. Allostatic emotional regulation in organizationally complex organisms permits enhanced adaptive behavioural flexibility relative to more reactive homeostatic dynamical systems. We discuss emotions as regulatory phenomena and provide a brief description of work in progress that will facilitate the gleaning of insights in this regard.
In this article, we review the nature of the functional and causal relationship between neurophysiologically/psychologically generated states of emotional feeling and action tendencies and extrapolate a novel perspective. Emotion theory, over the past century and beyond, has tended to regard feeling and action tendency as independent phenomena: attempts to outline the functional and causal relationship that exists between them have been framed therein. Classically, such relationships have been viewed as unidirectional, but an argument for bidirectionality rooted in a dynamic systems perspective has gained strength in recent years whereby the feeling-action tendency relationship is viewed as a composite whole. On the basis of our review of somatic-visceral theories of feelings, we argue that feelings are grounded upon neural-dynamic representations (elevated and stable activation patterns) of action tendency. Such representations amount to predictions updated by cognitive and bodily feedback. Specifically, we view emotional feelings as minimalist predictions of the action tendency (what the agent is physiologically and cognitively primed to do) in a given situation. The essence of this point is captured by our exposition of action tendency prediction-feedback loops with we consider, above all, in the context of emotion regulation, and in particular, of emotional regulation of goal-directed behavior. The perspective outlined may be of use to emotion theorists, computational modelers, and roboticists.
The somatic marker hypothesis (SMH) posits that the role of emotions and mental states in decision-making manifests through bodily responses to stimuli of import to the organism’s welfare. The Iowa Gambling Task (IGT), proposed by Bechara and Damasio in the mid-1990s, has provided the major source of empirical validation to the role of somatic markers in the service of flexible and cost-effective decision-making in humans. In recent years the IGT has been the subject of much criticism concerning: (1) whether measures of somatic markers reveal that they are important for decision-making as opposed to behaviour preparation; (2) the underlying neural substrate posited as critical to decision-making of the type relevant to the task; and (3) aspects of the methodological approach used, particularly on the canonical version of the task. In this paper, a cognitive robotics methodology is proposed to explore a dynamical systems approach as it applies to the neural computation of reward-based learning and issues concerning embodiment. This approach is particularly relevant in light of a strongly emerging alternative hypothesis to the SMH, the reversal learning hypothesis, which links, behaviourally and neurocomputationally, a number of more or less complex reward-based decision-making tasks, including the ‘A-not-B’ task – already subject to dynamical systems investigations with a focus on neural activation dynamics. It is also suggested that the cognitive robotics methodology may be used to extend systematically the IGT benchmark to more naturalised, but nevertheless controlled, settings that might better explore the extent to which the SMH, and somatic states per se, impact on complex decision-making.
With the present study we report the first application of a recently proposed model for realistic microbial fuel cells (MFCs) energy generation dynamics, suitable for robotic simulations with minimal and extremely limited computational overhead. A simulated agent was adapted in order to engage in a viable interaction with its environment. It achieved energy autonomy by maintaining viable levels of the critical variables of MFCs, namely cathodic hydration and anodic substrate biochemical energy. After unsupervised adaptation by genetic algorithm, these crucial variables modulate the behavioral dynamics expressed by viable robots in their interaction with the environment. The analysis of this physically rooted and self-organized dynamic action selection mechanism constitutes a novel practical contribution of this work. We also compare two different viable strategies, a self-organized continuous and a pulsed behavior, in order to foresee the possible cognitive implications of such biologicalmechatronics hybrid symbionts in a novel scenario of ecologically grounded energy and motivational autonomy.
The coupling between a body (in an extended sense that encompasses both neural and non-neural dynamics) and its environment is here conceived as a critical substrate for cognition. We propose and discuss the plan for a neurocomputational cognitive architecture for robotic agents, so far implemented in its minimalist form for supporting the behavior of a simple simulated agent. A non-neural internal bodily mechanism (crucially characterized by a time scale much slower than the normal sensory-motor interactions of the robot with its environment) extends the cognitive potential of a system composed of purely reactive parts with a dynamic action selection mechanism and the capacity to integrate information over time. The same non-neural mechanism is the foundation for a novel, minimalist anticipatory architecture, capable of swift re-adaptation to related yet novel tasks.
The coupling between a body (in an extended sense that encompasses both neural and non-neural dynamics) and its environment is here conceived as a critical substrate for cognition. We propose and discuss the plan for a neurocomputational cognitive architecture for robotic agents, so far implemented in its minimal form for supporting the behavior of a simple simulated robotic agent. A non-neural internal bodily mechanism (crucially characterized by a time scale much slower than the normal sensory-motor interactions of the robot with its environment) extends the cognitive potential of a system composed of purely reactive parts with a dynamic action selection mechanism and the capacity to integrate information over time. The same non-neural mechanism is the foundation for a novel, minimalist anticipatory architecture, implementing our bodily-anticipation hypothesis and capable of swift readaptation to related yet novel tasks.1
The new scenarios of contemporary adaptive robotics seem to suggest a transformation of the traditional methods. In the search for new approaches to the control of adaptive autonomous systems, the mind becomes a fundamental source of inspiration. In this paper we anticipate, through the use of simulation, the cognitive and behavioral properties that emerge from a recent prototype robotic platform, EcoBot, a family of bio-mechatronic symbionts provided with an 'artificial metabolism', that has been under physical development during recent years. Its energy reliance on a biological component and the consequent limitation of its supplied energy determine a special kind of dynamic coupling between the robot and its environment. Rather than just an obstacle, energetic constraints become the opportunity for the development of a rich set of behavioral and cognitive properties.
The coupling between a body (in an extended sense that encompasses both neural and non-neural dynamics) and its environment is here conceived as a critical substrate for cognition. We propose and discuss the plan for a neurocomputational cognitive architecture for robotic agents, so far implemented in its minimal form for supporting the behavior of a simple simulated robotic agent. A non-neural internal bodily mechanism (crucially characterized by a time scale much slower than the normal sensory-motor interactions of the robot with its environment) extends the cognitive potential of a system composed of purely reactive parts with a dynamic action selection mechanism and the capacity to integrate information over time. The same non-neural mechanism is the foundation for a novel, minimalist anticipatory architecture, implementing our bodily-anticipation hypothesis and capable of swift re-adaptation to related yet novel tasks.
Starting from the situated and embodied perspective on the study of cognition as a source of inspiration, this paper programmatically outlines a path towards an experimental exploration of the role of the body in a minimal anticipatory cognitive architecture. Cognition is here conceived and synthetically analyzed as a broadly extended and distributed dynamic process emerging from the interplay between a body, a nervous system and their environment. Firstly, we show how a non-neural internal state, crucially characterized by slowly changing dynamics, can modulate the activity of a simple neurocontroller. The result, emergent from the use of a standard evolutionary robotic simulation, is a selforganized, dynamic action selection mechanism, effectively operating in a context dependent way. Secondly, we show how these characteristics can be exploited by a novel minimalist anticipatory cognitive architecture. Rather than a direct causal connection between the anticipationprocess and the selection of the appropriate behavior, it implements a model for dynamic anticipation that operates via bodily mediation (bodily-anticipation hypothesis). This allows the system to swiftly scale up to more complex tasks never experienced before, achieving flexible and robust behavior with minimal adaptive cost.