How Morphological Computation Shapes Integrated Information in Embodied Agents

The Integrated Information Theory provides a quantitative approach to consciousness and can be applied to neural networks. An embodied agent controlled by such a network influences and is being influenced by its environment. This involves, on the one hand, morphological computation within goal directed action and, on the other hand, integrated information within the controller, the agent's brain. In this article, we combine different methods in order to examine the information flows among and within the body, the brain and the environment of an agent. This allows us to relate various information flows to each other. We test this framework in a simple experimental setup. There, we calculate the optimal policy for goal-directed behavior based on the “planning as inference” method, in which the information-geometric em-algorithm is used to optimize the likelihood of the goal. Morphological computation and integrated information are then calculated with respect to the optimal policies. Comparing the dynamics of these measures under changing morphological circumstances highlights the antagonistic relationship between these two concepts. The more morphological computation is involved, the less information integration within the brain is required. In order to determine the influence of the brain on the behavior of the agent it is necessary to additionally measure the information flow to and from the brain.

Morphological Computation in Plant Seeds for a New Generation of Self-Burial and Flying Soft Robots

Plants have evolved different mechanisms to disperse from parent plants and improve germination to sustain their survival. The study of seed dispersal mechanisms, with the related structural and functional characteristics, is an active research topic for ecology, plant diversity, climate change, as well as for its relevance for material science and engineering. The natural mechanisms of seed dispersal show a rich source of robust, highly adaptive, mass and energy efficient mechanisms for optimized passive flying, landing, crawling and drilling. The secret of seeds mobility is embodied in the structural features and anatomical characteristics of their tissues, which are designed to be selectively responsive to changes in the environmental conditions, and which make seeds one of the most fascinating examples of morphological computation in Nature. Particularly clever for their spatial mobility performance, are those seeds that use their morphology and structural characteristics to be carried by the wind and dispersed over great distances (i.e. “winged” and “parachute” seeds), and seeds able to move and penetrate in soil with a self-burial mechanism driven by their hygromorphic properties and morphological features. By looking at their motion mechanisms, new design principles can be extracted and used as inspiration for smart artificial systems endowed with embodied intelligence. This mini-review systematically collects, for the first time together, the morphological, structural, biomechanical and aerodynamic information from selected plant seeds relevant to take inspiration for engineering design of soft robots, and discusses potential future developments in the field across material science, plant biology, robotics and embodied intelligence.

Bodily Processing: What Progress Has Been Made in Understanding the Embodiment of Computing Systems?

" In this article I will address the issue of the embodiment of computing sys-tems from the point of view distinctive of the so-called Unconventional Computation, focusing on the paradigm known as Mor-phological Computation. As a first step, I will contextualize Morphological Computa-tion within the disciplinary field of Embod-ied Artificial Intelligence: broadly con-ceived, Embodied Artificial Intelligence may be characterized as embracing both conventional and unconventional ap-proaches to the artificial emulation of natu-ral intelligence. Morphological Computa-tion stands out from other paradigms of unconventional Embodied Artificial Intelli-gence in that it discloses a new, closer kind of connection between embodiment and computation. I will further my investigation by briefly reviewing the state-of-the-art in Morphological Computation: attention will be given to a very recent trend, whose core concept is that of “organic reconfigu-rability”. In this direction, as a final step, two advanced cases of study of organic or living morphological computers will be pre-sented and discussed. The prospect is to shed some light on our title question: what progress has been made in understanding the embodiment of computing systems? Keywords: Embodied Artificial Intelligence; Morphological Computation; Reservoir Compu-ting Systems; Organic Reconfigurability; 3D Bio-Printed Synthetic Corneas; Xenobots "

Online Morphological Adaptation for Tactile Sensing Augmentation

Sensor morphology and structure has the ability to significantly aid and improve tactile sensing capabilities, through mechanisms such as improved sensitivity or morphological computation. However, different tactile tasks require different morphologies posing a challenge as to how to best design sensors, and also how to enable sensor morphology to be varied. We introduce a jamming filter which, when placed over a tactile sensor, allows the filter to be shaped and molded online, thus varying the sensor structure. We demonstrate how this is beneficial for sensory tasks analyzing how the change in sensor structure varies the information that is gained using the sensor. Moreover, we show that appropriate morphology can significantly influence discrimination, and observe how the selection of an appropriate filter can increase the object classification accuracy when using standard classifiers by up to 28%.

Tactile sensing biohybrid soft E-skin based on bioimpedance using aloe vera pulp tissues

Soft and flexible E-skin advances are a subset of soft robotics field where the soft morphology of human skin is mimicked. The number of prototypes that conformed the use of biological tissues within the structure of soft robots—to develop “Biohybrid Soft Robots”—has increased in the last decade. However, no research was conducted to realize Biohybrid E-skin. In this paper, a novel biohybrid E-skin that provides tactile sensing is developed. The biohybrid E-skin highly mimics the human skin softness and morphology and can sense forces as low as 0.01 newton . The tactile sensing feature is augmented through the use of Aloe Vera pulp, embedded in underlying channel, where the change in its bioimpedance is related to the amount of force exerted on the E-skin surface. The biohybrid E-skin employs high biomimicry as the sensorial output is an oscillating signal similar to signals sent from the human sensing neurons to the brain. After investigating different channel geometries, types of filling tissues, and usage of two silicone materials, their frequency-force behaviour is modelled mathematically. Finally, a functional multichannel prototype “ImpEdded Skin” is developed. This prototype could efficiently detect the position of a tactile touch. This work employs the development of discrete sensing system that exhibits morphological computation that consequently enhances performance.

Control vs. Constraint: Understanding the Mechanisms of Vibration Transmission During Material-Bound Information Transfer

Material-bound vibrations are ubiquitous in the environment and are widely used as an information source by animals, whether they are generated by biotic or abiotic sources. The process of vibration information transfer is subject to a wide range of physical constraints, especially during the vibration transmission phase. This is because vibrations must travel through materials in the environment and body of the animal before reaching embedded mechanosensors. Morphology therefore plays a key and often overlooked role in shaping information flow. Web-building spiders are ideal organisms for studying vibration information transfer due to the level of control they have over morphological traits, both within the web (environment) and body, which can give insights for bioinspired design. Here we investigate the mechanisms governing vibration information transfer, including the relative roles of constraints and control mechanisms. We review the known and theoretical contributions of morphological and behavioral traits to vibration transmission in these spiders, and propose an interdisciplinary framework for considering the effects of these traits from a biomechanical perspective. Whereas morphological traits act as a series of springs, dampers and masses arranged in a specific geometry to influence vibration transmission, behavioral traits influence these morphologies often over small timescales in response to changing conditions. We then explore the relative roles of constraints and control mechanisms in shaping the variation of these traits at various taxonomic levels. This analysis reveals the importance of morphology modification to gain control over vibration transmission to mitigate constraints and essentially promote information transfer. In particular, we hypothesize that morphological computation is used by spiders during vibration information transfer to reduce the amount of processing required by the central nervous system (CNS); a hypothesis that can be tested experimentally in the future. We can take inspiration from how spiders control vibration transmission and apply these insights to bioinspired engineering. In particular, the role of morphological computation for vibration control could open up potential developments for soft robots, which could use multi-scale vibration sensory systems inspired by spiders to quickly and efficiently adapt to changing environments.