Towards modelling affect and emotions in autonomous agents with recurrent fuzzy systems

Abstract

In this paper, we present a computational model of the elicitation and the dynamics of affective states in autonomous agents based on fuzzy methodology. Fuzzy logic has been used in computational modelling of affect for 15 years, but has not yet been employed to bridge the particular benefits of fuzzy modelling with dynamical systems approaches to understanding emotions. The information-processing architecture scheme we propose for the calculations which are involved in affective processes can be instantiated in a distributed fashion. This enables 'programming' of the agent's affective interpretation of its relationship with its environment with respect to different objects and their behavior towards the agent, as well as its own internal motivational states. In our simulations, we emphasize dynamic properties of affective processes and behaviors resulting therefrom. In addition, we sketch theoretical ideas of internal effects of different affective states that we are currently implementing.

Is It Necessary to Integrate Evo-Devo to the Analysis and Construction of Artificial Emotional Systems?

“…According to these authors, for an artificial agent to have emotions sensu stricto it is necessary that a mechanism linked to its bodily states guides its actions, its decisions, and allows it to achieve better performance within its environment. In this context, it is claimed that if these artificial agents interact in the world through internal mechanisms—for example, homeostasis—that regulate their behavior, a better understanding of the role of the underlying mechanisms of emotions in the study of cognition will be achieved (see, for instance, Coutinho et al, 2005 ; Parisi and Petrosino, 2010 ; Schneider and Adamy, 2014 ). Nevertheless, the absence of an evo-devo approach in these two subjects is noteworthy.…”

“…The ideas of Ziemke ( 2016 ) are of additional interest for their emphasis on the “hard problem” status that characterizes the analysis and construction of AESs. Here we refer to implementations (of an abstract nature in standard, non-embodied cognitive science/AI research; Coutinho et al, 2005 ; Parisi and Petrosino, 2010 ; Schneider and Adamy, 2014 ) in which “states (lead to) increase the correctness and effectiveness of the motivational decisions (of the system) by influencing the current intensity of the different motivations” (modified from Parisi and Petrosino, 2010 , p. 3). We consider that, as a matter of principle and in ways that remain to be empirically studied in detail under increasingly stringent methodological criteria, most elements of the (eco-)evo-devo/EES conceptual framework (lower ellipse, Figure 1 ) are applicable to the study of AESs.…”

Robot motion control based on anticipatory emotions

“…Schneider and Adamy present an agent in a grid world making decisions to extend its lifespan, given limited available energy plus the negative effect of predators. 17 More advanced proposals to improve adaptation in robots require hierarchical architectures, where emulated emotions and emulated reasoning work together to elicit an appropriate behaviour, or an entire navigation plan. For instance, Goerke defines three levels of processing.…”

Computational Approaches to Modeling Artificial Emotion – An Overview of the Proposed Solutions

“…(Dias, 2005), Thespian (Mei et al, 2006), Peactidm (Marinier et al, 2009), Wasabi (Becker-Asano, 2008, AR (Elliott, 1992), CyberCafe (Rousseau, 1996), Silas (Blumberg, 1996), Cathexis (Velásquez and Maes, 1997), OZ (Reilly and Bates, 1992), MAMID (Hudlicka, 2005), CRIBB (Davis and Lewis, 2003), and Affect and Emotions (Schneider and Adamy, 2014). However, some of them are either freely/weakly connected with the psychological aspects of emotions (e.g., ACRES) or away from a strict mathematical wording of the issue (FAtiMA Modular, MAMID).…”