Biomimetic and Biohybrid Systems

To better understand the role of natural dynamics in motor control, we have constructed a mathematical model of crawling mechanics in larval

Drosophila

.

The model accounts for key anatomical features such as a segmentally patterned, viscoelastic outer body wall (cuticle); a non-segmented inner cavity (haemocoel) filled with incompressible fluid that enables visceral pistoning; and claw-like protrusions (denticle bands) giving rise to asymmetric friction.

Under conditions of light damping and low forward kinetic friction, and with a single cuticle segment initially compressed, the passive dynamics of this model produce wave-like motion resembling that of real larvae. The presence of a volume-conserving hydrostatic skeleton allows a wave reaching the anterior of the body to initiate a new wave at the posterior, thus recycling energy. Forcing our model with a sinusoidal input reveals conditions under which power transfer from control to body may be maximised. A minimal control scheme using segmentally localised positive feedback is able to exploit these conditions in order to maintain wave-like motion indefinitely. These principles could form the basis of a design for a novel, soft-bodied, crawling robot.

We present a novel soft limb quadruped robot “FASTT,” with a simple and cheap design of its legs for dynamic locomotion aimed to expand the applications of soft robotics in mobile robots. The pneumatically actuated soft legs are self-stabilizing, adaptive to ground, and have variable stiffness, all of which are essential properties of locomotion that are also found in biological systems. We tested the soft legs for the pace, trot, and gallop gait and found them to move with a forward velocity for each gait with robustness. The legs were able to produce a flight and stance phase as a result of the body-environment interaction and also support the weight of the body while two legs were in flight phase and two in stance phase. The soft robot also exhibited two different postures i.e. sprawl and semi-erect which can also be found in some biological species as the crocodile. Moreover, the robot is safe to interact with. The results highlight the effectiveness of the soft limbs to produce dynamic locomotion which provides potential for application in uncertain environments.

In order to mimic and better understand the way an earthworm uses its many segments to navigate diverse terrain, this paper describes the design, performance, and sensing capabilities of a new modular soft robotic worm. The robot, Compliant Modular Mesh Worm (CMMWorm), utilizes a compliant mesh actuated at modular segments to create waveforms along its body. These waveforms can generate peristaltic motion of the body similar to that of an earthworm. The modular mesh is constructed from 3-D printed and commercially available parts allowing for the testing of a variety of components that can be easily interchanged. In addition to having independently controlled segments and interchangeable mesh properties, CMMWorm also has greater range of contraction (52% of maximum diameter) than our previous robot Softworm (73% of maximum diameter). The six-segment robot can traverse flat ground and pipes. We show that a segment is able to detect the wall of a pipe and return to its initial position using actuator-based load-sensing. A simple kinematic model predicts the outer diameter of the worm robot’s mesh as a function of encoder position.

WormTIP is a novel lightweight self-actuating exploratory sensor, using a pneumostatic vessel and a dielectric elastomeric actuator (DEA) to create an active sensory tip capable of object shape determination as part of a flexible soft robot. Utilising the coupling of a static fluid vessel, the DEA is paired with a sensory membrane with internal papillae mimicking the internal morphology found in the fingertip. The sensory membrane is extended onto an object, conforming to its surface. Experimental results are presented which show the detection of shapes using particle velocimetry and papillae density analysis. These are preliminary results which show the potential of the WormTIP, which is the focus of ongoing work. The device is aimed for use as a self-contained palpating sensor, or as an attachment to a bio-inspired robotic worm forming a self-contained exploratory vehicle with the device acting as the sensory appendage or proboscis.

Mimicking biological system successfully requires that the materials used in building such a system are qualitatively similar to that offered by the biological systems. One of such material is carbon filled natural rubber. Furthermore, biological systems implements various forms of support structures of which the ones imitated in this work is referred to as muscular hydrostatic support as opposed to fluid filled hydrostatic support. A muscular hydrostatic model proposed could be adapted to 3D motion but a planar joint/support was implemented as a proof of concept based on

teleost

fish - 394.01 mm long Mackerel. Static test indicate a well mimicked tail motion even with just three actuators. Turning test of the robotic fish inside tight box was successful as it was able to turn after several attempts. Also the robot was able to swim in a shallow pool of water where it attained 0.985m/s linear speed.

Development of soft robotic devices with grasping capabilities is an active research area. The inherent property of soft materials, to distribute contact forces, results in a more effective robot/environment interaction with simpler control. In this paper, a three-finger under-actuated adaptable soft gripper is proposed, highlighting the design and manufacturing process. A novel design and actuation principle have been implemented to obtain the desired grasping abilities, from mechanical properties of materials and structures. Soft materials have been used to make each finger, for a high adaptability of the gripper to different shapes. We implemented an under-actuated mechanism through a wire loop actuation system, that helps achieving passive adaptation during grasping. Passive adaptability allows to drive the device with a reduced number of control parameters. The soft gripper has been lodged into an experimental setup endowed with one actuation unit for the synchronous flexion of its fingers. Grasping and holding capabilities have been tested by evaluating the grasp stability with target objects varying in shape, size and material. Adaptability makes this soft device a good application of morphological computation principles in bio-inspired robots design, where proper design of mechanical features simplifies control.

A novel technique for determining the relative visco-density of fluids using an actuated flexible beam inspired by the tactile whiskers of marine mammals is presented. This was developed for the in-situ calibration of a tactile whisker based system for measuring flow velocity around autonomous robots working in complex underwater environments.

A new thermoplastic protein complex, Squid Ring Teeth (SRT), has been adapted for use in the artificial reconstruction of a

Manduca sexta

wing. The SRT protein complex exhibits consistent material properties over a wide range of temperatures (25°C to 196°C) and retains it mechanical integrity across a large frequency spectrum (0.1 Hz to 150 Hz). Insect-inspired wings comprised of SRT can therefore be reliable and robust, which are essential characteristics for flapping wing MAVs (FWMAV). The preliminary results in this paper suggest that a thorough analysis of an SRT-based wing be conducted using load cell, optical digitization, and PIV techniques. With these results, we believe it will be possible to accurately mimic the

M. sexta

wing in order to pave the way for next generation FWMAV development.

The study of biological organisms may aid with designing more dynamic, adaptable robots. In this paper, we quantitatively studied the coupling of kinematics and kinetics in the common earthworm,

Lumbricus terrestris

. Our data correlates changes in worm segment shape to variable, non-uniform load distribution of worm weight. This presumably leads to variable friction forces. Understanding the way the worm exerts these forces may help us implement peristalsis in robots in diverse real-world environments. In our preliminary data, at the front of the worm, the segments with the widest diameter bear the most weight and anchor the worm to the ground during motion, as we hypothesized. The rear segments also exhibit variation in ground reaction forces. However, for rear segments, the peak kinetic waves are phase-shifted from the kinematic waves. Future work will explore this phenomenon.

Frogs can swim adaptively in the water dexterously utilizing interactions between body biomechanics and fluid dynamics.

We have been developing an aquatic frog robot, Xenopus-noid, which has similar musculoskeletal structure as its biological counterpart,

Xenopus laevis

. This robot allows us to study the interaction between the biomechanical structure of the frog and the fluid dynamics during swimming locomotion in a natural context. In this report, particle image velocimetry (PIV) is used for visualizing wakes generated by the Xenopus-noid. Experimental results demonstrate that the robot can produce appropriate wakes for swimming if we implement a rigid beam that mimics the function of the Semimembranosus (SM) muscle in

Xenopus laevis

. The function is utilized for proper posture, that is to say, this muscle prevents a hyperextension of the knee.

Modern architectural designs create dynamic and flexible spaces, able to adapt to the ever-changing environment by virtue of temporary and convertible structures. Biomimetics is the applied science that, through the imitation of Nature, finds the solution to a human problem. The unfolding of a beetle wing and the blooming of a swirl flower were recognised as having outstanding features to be mimicked for the creation of deployable canopies. This paper focuses on the analysis methodology of the two biomimetic, deployable structures with multiple degrees of freedom. The general validity of a pseudo-static analysis was proved based on time-stepping the geometry at set deployment stages with optimisation of multiple, potential deployment sequences.

Here we present a tactile sensing capsule endoscopy system. Whilst current capsule endoscopy utilises cameras to diagnose lesions on the surface of the gastrointestinal tract lumen, this proposal uses remote palpation to stimulate a bio-inspired tactile sensing surface that deforms under the impression of hard raised objects. This provides the capability to characterise tissue density and lesions more deeply than the lumen surface.

Birds’ feathers are equipped with hook-like structures called friction barbules, which prevent separation and rubbing between feathers under nominal flow conditions. This paper presents a segmented wing prototype that uses controllable dry adhesives to mimic the function of friction barbules. The adhesives latch wing segments together during moderate flight conditions and allow them to separate in extreme conditions. We present the characteristics of adhesive patches and their performance as they are incorporated into a flexible wing prototype. The attachment force is a function of the applied shear stress. We then present results of a wind tunnel test to evaluate the aerodynamic effect of gaps formed as wing segments unlatch and separate. The separation of wing segments delays stall and reduces overall drag, which could improve the ability of an unmanned air vehicle to fly in gusty conditions.

An autonomous vehicle controlled using real-time obstacle-avoidance algorithms and ultrasound was constructed to understand the active sensing system of the bat. The vehicle was designed to mimic bat behavior in which 1) the outgoing pulse was emitted toward the obstacle detected by the previous echo (obstacle aiming) and 2) the interpulse interval was adjusted using the distance to the detected object. As a result, the obstacle-aiming system facilitated obstacle avoidance by keeping the obstacle in the center of the beam sight of the vehicle.

Behavioral experiments involving a bat avoiding obstacles demonstrated that the bat responds to multiple echoes and then decides the direction of the next outgoing pulse. Based on this behavior, a multi-object-detection navigation system was proposed to enable a vehicle to move in more complicated space that it failed to navigate previously. Our findings suggest that the bat behavioral strategies provide new perspectives for engineering involving simple sensing.

Miniaturized artificial hearing organ with excellent sensitivity and wide dynamic frequency range over human hearing range, while requiring small amount of energy, is important step to develop artificial systems interacting in human living space. This paper presents the development of piezoelectric artificial cochlea (PAC) capable of analyzing incoming vibratory signals over human hearing range without external power source. The design, component and function of PAC were inspired by those of human cochlea. The PAC was made of corona-poled piezoelectric thin film with vibrating membrane part of unique shape. The vibration displacement of membrane was measured using laser Doppler vibrometer and analyzed to show the frequency separation of the developed PAC. The experimental results of mechanical vibratory behavior demonstrated successful separation of incoming signals into 13 different frequency bands depending on their frequency over 300 Hz ~ 6,000 Hz.

In this study, a tiny artificial compound eye (diameter 15mm) named CurvACE (which stands for Curved Artificial Compound Eye), was endowed with hyperacuity, based on an active visual process inspired by the retinal micro-movements occurring in the fly’s compound eye. A periodic (1-D, 50-Hz) micro-scanning movement with a range of a few degrees (

$$5^\circ $$

) enables the active CurvACE to locate contrasting objects with a 40-fold greater accuracy which was restricted by the narrow interommatidial angle of about

$$4.2^\circ $$

. This local hyperacuity was extended to a large number of adjacent ommatidia in a novel visual processing algorithm, which merges the output signals of the local processing units running in parallel on a tiny, cheap micro-controller requiring very few computational resources. Tests performed in a textured (indoor) or natural (outdoor) environment showed that the active compound eye serves as a contactless angular position sensing device, which is able to assess its angular position relative to the visual environment. As a consequence, the vibrating compound eye is able to measure very low rotational optic flow up to

$$ 20^\circ /s $$

and perform a short range odometry knowing the altitude, which are two tasks of great interest for robotic applications.

In this paper, we describe the implementation of a closed-loop control architecture on a bio-hybrid robotic system. The control loop uses the spiking activity from two motion-sensitive H1-cells recorded in both halves of the blowfly’s brain as visual feedback signals that are sent to an ARM processor, programmed to establish a brain machine interface. The resulting output controls the movements of the robot which, in turn, generates optic flow that modifies the activity of the H1-cells. Instead of being inhibited by front-to-back optic flow would the robot move forward in a straight line, the closed-loop system autonomously produces an oscillatory trajectory, alternatingly stimulating both H1-cells with back-to-front optic flow. The spike rate information of each cell is then used to control the speed of each robot wheel, on average driving the robot in the forward direction. Our extracellular recordings from the two cells show similar spike rate oscillation frequencies and amplitude, but opposite phases. From our experiments we derive parameters relevant for the future implementation of collision avoidance capabilities. Finally, we discuss a control algorithm that combines positive and negative feedback to drive the robot.

We present Mantisbot, a 28 degree of freedom robot controlled in real-time by a neural simulation. MantisBot was designed as a 13.3:1 model of a male Tenodera sinensis with the animal’s predominant degrees of freedom. The purpose of this robot is to investigate two main topics: 1. the control of targeted motion, such as prey-directed pivots and striking, and 2. the role of descending commands in transitioning between behaviors, such as standing, prey stalking, and walking. In order to more directly use data from the animal, the robot mimics its kinematics and range of motion as closely as possible, uses strain gages on its legs to measure femoral strain like insects, and is controlled by a realistic neural simulation of networks in the thoracic ganglia. This paper summarizes the mechanical, electrical, and software design of the robot, and how its neural control system generates reflexes observed in insects. It also presents preliminary results; the robot is capable of supporting its weight on four or six legs, and using sensory information for adaptive and corrective reflexes.

Bio-inspired guidance principles involving no reference frame are presented here and were implemented in a rotorcraft called Beerotor, which was equipped with a minimalistic panoramic optic flow sensor and no accelerometer, no inertial measurement unit (IMU) [

9

], as observed in flying insects (The halters of Diptera are only sensitive to rotation rates). In the present paper, the vertical optic flow was used as an additional cue whereas the previously published Beerotor’s visuo-motor systems only used translational optic flow cues [

9

]. To test these guidance principles, we built a tethered tandem rotorcraft called Beerotor (80g), which flies along a high-roofed tunnel. The aerial robot adjusts its pitch and hence its speed, hugs the ground and lands safely without any need for an inertial reference frame. The rotorcraft’s altitude and forward speed are adjusted via several optic flow feedback loops piloting respectively the lift and the pitch angle on the basis of the common-mode and differential rotor speeds, respectively as well as an active system of reorientation of a quasi-panoramic eye which constantly realigns its gaze, keeping it parallel to the nearest surface followed. Safe automatic terrain following and landing were obtained with the active eye-reorientation system over rugged terrain, without any need for an inertial reference frame.

Desert ants are expert navigators, foraging over large distances using visually guided routes. Recent models of route following can reproduce aspects of route guidance, yet the underlying motor patterns do not reflect those of foraging ants. Specifically, these models select the direction of movement by rotating to find the most familiar view. Yet scanning patterns are only occasionally observed in ants. We propose a novel route following strategy inspired by klinokinesis. By using familiarity of the view to modulate the magnitude of alternating left and right turns, and the size of forward steps, this strategy is able to continually correct the heading of a simulated ant to maintain its course along a route. Route following by klinokinesis and visual compass are evaluated against real ant routes in a simulation study and on a mobile robot in the real ant habitat. We report that in unfamiliar surroundings the proposed method can also generate ant-like scanning behaviours.

A baseline model for testing how afferent muscle feedback affects both timing and activation levels of muscle contractions has been constructed. We present an improved version of the neuromechanical model from our previous work [

6

]. This updated model has carefully tuned muscles, feedback pathways, and central pattern generators (CPGs). Kinematics and force plate data from trotting rats were used to better design muscles for the legs. A recent pattern generator topology [

15

] is implemented to better mimic the rhythm generation and pattern formation networks in the animal. Phase-space and numerical phase response analyses reveal the dynamics underlying CPG behavior, resulting in an oscillator that produces both robust cycles and favorable perturbation responses. Training methods were used to tune synapse properties to shape desired motor neuron activation patterns. The result is a model which is capable of self-propelled hind leg stepping and will serve as a baseline as we investigate the effects changes in afferent feedback have on muscle activation patterns.

Here we used multi-electrode array technology to examine the encoding of temporal information in dissociated hippocampal networks. We demonstrate that two connected populations of neurons can be trained to encode a defined time interval, and this memory trace persists for several hours. We also investigate whether the spontaneous firing activity of a trained network, can act as a template for copying the encoded time interval to a naïve network. Such findings are of general significance for understanding fundamental principles of information storage and replication.

A muscle cell actuator has been attracting a lot of attention since it is a key technology for realizing bio-machine hybrid systems. This study especially intend to deal with micro robot driven by real muscle cell actuators. To fabricate the actuator, we should study how to control cell aggregation for efficient power generation. This paper proposes a method for remodeling muscle cells by exploiting mechanical stimulus so that we can get an appropriate structure of the actuator. The experimental results demonstrate that the three factors, cell density, cell-matrix adhesion, and mechanical stimulation period, largely contribute to the remodeling of the muscle cells.

We developed a biologically plausible control algorithm to move the eyes of a six degrees of freedom robotic head in a human-like manner. Our neurocontroller, written with the neural simulator

Nengo

, integrates different biological neural models of eye movements, such as microsaccades, saccades, vestibular-ocular reflex, smooth pursuit and vergence. The coordination of the movements depends on the stream of sensory information acquired by two silicon retinas used as eyes and by an inertial measurement unit, which serves as a vestibular system. The eye movements generated by our neurocontroller resemble those of humans when exposed to the same visual input. This robotic platform can be used to investigate the most efficient exploration strategies used to extract salient features from either a static or dynamic visual scene. Future research should focus on technical enhancements and model refinements of the system.

Recently, the concept of a ‘companion robot’ as a healthcare tool has been popularised, and even commercialised. We present MIRO, a robot that is biomimetic in aesthetics, morphology, behaviour, and control architecture. In this paper, we review how these design choices affect its suitability for a companionship role. In particular, we consider how emulation of the familiar body language and other emotional expressions of mammals may facilitate effective communication with naïve users through the reliable evocation of intended perceptions of emotional state and intent. We go on to present a brief pilot study addressing the question of whether shared cultural signals can be relied upon, similarly, as components of communication systems for companion robots. Such studies form part of our ongoing effort to understand and quantify human responses to robot expressive behaviour and, thereby, develop a methodology for optimising the design of social robots by accounting for individual and cultural differences.

In order to follow a moving visual target, humans generate voluntary smooth pursuit eye movements. The purpose of smooth pursuit eye movements is to minimize the retinal slip, i.e. the target velocity projected onto the retina. In this paper we propose a model able to integrate the major characteristics of visually guided and predictive control of the smooth pursuit. The model is composed of an Inverse Dynamics Controller (IDC) for the feedback control, a neural predictor for the anticipation of the target motion and a Weighted Sum module that is able to combine the previous systems in a proper way. In order to validate the general model, two implementations with two different IDC controllers have been carried out. The first one uses a backstepping-based controller to generate velocity motor commands for the eye movements and the other one uses a bio-inspired neurocontroller to generate position motor commands for eye-neck coordinated movements. Our results, tested on the iCub robot simulator, show that both implementations can use prediction for a zero-lag visual tracking, a feedback based control for “unpredictable” target pursuit and can combine these two approaches by properly switching from one to the other, guaranteeing a stable visual pursuit.

Although deep neural networks (DNNs) have demonstrated impressive results during the last decade, they remain highly specialized tools, which are trained – often from scratch – to solve each particular task. The human brain, in contrast, significantly re-uses existing capacities when learning to solve new tasks. In the current study we explore a block-modular architecture for DNNs, which allows parts of the existing network to be re-used to solve a new task without a decrease in performance when solving the original task. We show that networks with such architectures can outperform networks trained from scratch, or perform comparably, while having to learn nearly 10 times fewer weights than the networks trained from scratch.

Autobiographical memory (AM) refers to the organisation of one’s experience into a coherent narrative. The exact neural mechanisms responsible for the manifestation of AM in humans are unknown. On the other hand, the field of psychology has provided us with useful understanding about the

functionality

of a bio-inspired synthetic AM (SAM) system, in a higher level of description. This paper is concerned with a top-down approach to SAM, where known components and organisation guide the architecture but the unknown details of each module are abstracted. By using Bayesian latent variable models we obtain a transparent SAM system with which we can interact in a structured way. This allows us to reveal the properties of specific sub-modules and map them to functionality observed in biological systems. The top-down approach can cope well with the high performance requirements of a bio-inspired cognitive system. This is demonstrated in experiments using faces data.

Crowdsourcing is a popular technique for distributing tasks to a group of anonymous workers over the web. Similarly, crowdseeding is any mechanism that extracts knowledge from the crowd, and then uses that knowledge to guide an automated process. Here we demonstrate a method that automatically distills features from a set of robot body plans designed by the crowd, and then uses those features to guide the automated design of robot body plans and controllers. This approach outperforms past work in which one feature was detected and distilled manually. This provides evidence that the crowd collectively possesses intuitions about the biomechanical advantages of certain body plans; we hypothesize that these intuitions derive from their experiences with biological organisms.

Procedural and Declarative knowledge play a key role in cognitive architectures for robots. These types of architectures use the human brain as inspiration to design control structures that allow robots to be fully autonomous, in the sense that their development depends only on their own experience in the environment. The two main components that make up cognitive architectures are models (prediction) and action-selection structures (decision). Models represent the declarative knowledge the robot acquires during its lifetime. On the other hand, action-selection structures represent the procedural knowledge, and its autonomous acquisition depends on the quality of the models that are being learned concurrently. The coupled learning of models and action-selection structures is a key aspect in robot development, and it has been rarely studied in the field. This work aims to start filling this gap by analyzing how these concurrent learning processes affect each other using an evolutionary-based cognitive architecture, the Multilevel Darwinist Brain, in a simulated robotic experiment

How experienced emotional states, induced by the events that emerge in our context, influence our behaviour? Are they an obstacle or a helpful assistant for our reasoning process? Antonio Damasio gave exhaustive answers to these questions through his studies on patients with brain injuries. He demonstrated how the emotions guide decision-making and he has identified a region of the brain which has a fundamental role in this process. Antoine Bechara devised a test to validate the proper functioning of that cortical region of the brain. Inspired from Damasio’s theories we developed a mechanism in an artificial agent that enables it to represent emotional states and to exploit them for biasing its decisions. We also implement the card gambling task that Bechara used on his patients as a validating test. Finally we put our artificial agent through this test for 100 trials. The results of this experiment are analysed and discussed highlighting the demonstrated efficiency of the implemented somatic marker mechanism and the potential impact of this system in the field of social robotics.

Learning processes which take place during the development of a biological nervous system enable it to extract mappings between external stimuli and its internal state. Precise egomotion estimation is essential to keep these external and internal cues coherent given the rich multisensory environment. In this paper we present a learning model which, given various sensory inputs, converges to a state providing a coherent representation of the sensory space and the cross-sensory relations. The developed model, implemented for 3D egomotion estimation on a quadrotor, provides precise estimates for roll, pitch and yaw angles.

Using a skill learning architecture being developed for the humanoid iCub, in this article we show through experiments how a cognitive robot can learn to imitate quickly the use of a tool after a teacher’s demonstration by recycling previously learnt knowledge from drawing experiments. The employed architecture incorporates novel principles for constructing a growing motor vocabulary in cognitive robots enabling “cumulative” and “swift” learning of skills through integration of multiple streams of learning like imitation and mental simulation. A central notion emphasized in the architecture is that movements can be represented in the form of ‘shapes’ instead of trajectories per se in order to liberate them from task specific details. The idea is to abstract out critical features/knowledge in the movement trajectory, which can later be reused in a context-independent manner.

Models of children’s cognitive development indicate that as children grow, they transition from using behavioral cues to knowledge of biology to determine a target’s animacy. This paper explores the impact of children’s’ ages and a humanoid robot’s expressive behavior on their perceptions of the robot, using a simple, low-demand measure. Results indicate that children’s ages have influence on their perceptions in terms of the robot’s status being a person, a machine, or a composite. Younger children (aged 6) tended to rate the robot as being like a person to a substantially greater extent than older children (aged 7) did. However, additional facially-expressive cues from the robot did not substantively impact on children’s responses. Implications for future HRI studies are discussed.

Human-Robot-Interaction (HRI) research is often built around the premise that the robot is serving to assist a human in achieving a human-led goal or shared task. However, there are many circumstances during HRI in which a robot may need the assistance of a human in shared tasks or to achieve goals. We use the ROBO-GUIDE model as a case study, and insights from social psychology, to examine two factors of user trust and situational ambiguity which may impact promote human user assistance towards a robot. These factors are argued to determine the likelihood of human assistance arriving, individuals’ perceived competence of the robot, and individuals’ trust towards the robot. We outline an experimental approach to test these proposals.

Sensory augmentation is one of the most exciting domains for research in human-machine biohybridicity. The current paper presents the design of a 2

nd

generation vibrotactile helmet as a sensory augmentation prototype that is being developed to help users to navigate in low visibility environments. The paper outlines a study in which the user navigates along a virtual wall whilst the position and orientation of the user’s head is tracked by a motion capture system. Vibrotactile feedback is presented according to the user’s distance from the virtual wall and their head orientation. The research builds on our previous work by developing a simplified “tactile language” for communicating navigation commands. A key goal is to identify language tokens suitable to a head-mounted tactile interface that are maximally informative, minimize information overload, intuitive, and that have the potential to become ‘experientially transparent’

We present an energetically autonomous robotic tadpole that uses a single membrane component for both electrical energy generation and propulsive actuation. The coupling of this small bio-inspired power source to a bio-inspired actuator demonstrates the first generation design for an energetically autonomous swimming robot consisting of a single membrane. An ionic polymer metal composite (IPMC) with a Nafion polymer layer is demonstrated in a novel application as the ion exchange membrane and anode and cathode electrode of a microbial fuel cell (MFC), whilst being used concurrently as an artificial muscle tail. In contrast to previous work using stacked units for increased voltage, a single MFC with novel, 0.88ml anode chamber architecture is used to generate suitable voltages for driving artificial muscle actuation, with minimal step up. This shows the potential of the small forces generated by IPMCs for propulsion of a bio-energy source. The work demonstrates great potential for reducing the mass and complexity of bio-inspired autonomous robots. The performance of the IPMC as an ion exchange membrane is compared to two conventional ion exchange membranes, Nafion and cation exchange membrane (CEM). The MFC anode and cathode show increased resistance following inclusion within the MFC environment.

Group-living animals often exhibit complex collective behaviors that emerge through the non-linear dynamics of social interactions between individuals. Previous studies have shown that it is possible to influence the collective decision-making process of groups of insects by integrating them with autonomous multi-robot systems. However, generating robot controller models for this particular task can be challenging. The main difficulties lie in accommodating group collective dynamics (macroscopic level) and agent-based models implemented in every individual robot (microscopic level). In this study, we show how such systems can be appropriately modeled, and how to use them to modulate the collective decision-making of cockroaches in a shelter-selection problem. We address two questions in this paper: first, how to optimize a microscopic model of cockroach behavior to exhibit the same collective behavior as a macroscopic model from the literature, and second, how to optimize the model describing robot behavior to modulate the collective behavior of the group of cockroaches.

In order for robots to be part of the education field, it is necessary to take into consideration the perception students have of them and of education in general. The aim of this study is to assess whether the role a robot plays in a classroom affects knowledge retrieval, subjective experience, and the perception of the learners. To investigate this, we developed an educational scenario and three questionnaires. The results show significant differences in the way the subjects perceived the robot as a tutor.

Dolphins identify objects using their sonar, which works by emitting short acoustic pulses with high bandwidth and high intensity. These echolocation impulses have a double chirp structure. The complex signal structure allows animals to collect more information than simply the distance to the object. They can evaluate object’s size, shape, and even the innards of the object, by processing the whole echo from the object.

The study of the dolphins’ clicks inspired a simulation of the signals for echolocation purposes. They are already used for object characterisation.

In addition, dolphins’ clicks are adaptive signals. Dolphins can change some parameters of the clicks during recognition process, which allows them to achieve better results for object characterisation.

This paper presents background and the main concept of the adaptive echolocation using bio-inspired signals. Implementation of adaptive echolocation is a new approach and can improve object characterisation and will help to achieve more accurate results.

TCERA (Tunable Compliance Energy Return Actuator) is a robotic emulation of a human femur-tibia system in a walking or running gait. TCERA features two parallel air springs to compliantly transmit or absorb torque about the knee. The radial positions of the springs as well as their relative lengths are adjustable, which effectively alter the torsional stiffness and equilibrium angle of the knee system, respectively. These functions are operable independently, and thus decouple the kinematic relationship and control of the torsional stiffness and femur-tibia angle. The actuator derives its inspiration from an analysis of the human gait, in which it can be seen that the torsional compliance about the knee is varied to specific stiffness values throughout the cycle. Special mechanical considerations have been taken into account to achieve swift changes in position without compromising the static integrity of the actuator.

We present an investigation on active control for intelligent object exploration using touch with a robotic hand. First, uncertainty from the exploration is reduced by a probabilistic method based on the accumulation of evidence through the interaction with an object of interest. Second, an intrinsic motivation approach allows the robot hand to perform intelligent active control of movements to explore interesting locations of the object. Passive and active perception and exploration were implemented in simulated and real environments to compare their benefits in accuracy and reaction time. The validation of the proposed method were performed with an object recognition task, using a robotic platform composed by a three-fingered robotic hand and a robot table. The results demonstrate that our method permits the robotic hand to achieve high accuracy for object recognition with low impact on the reaction time required to perform the task. These benefits make our method suitable for perception and exploration in autonomous robotics.

Based on the need for small scale compliant devices for use in medical robotics, there is increasing interest in the development of muscle actuated biobots. Such biobots are traditionally fabricated using nondegradable, synthetic polymers; however, such substrates require micro-patterning and additional treatments in order to promote cellular adhesion and induce cellular alignment. By using an organic substrate, such steps can be eliminated, and degradable scaffolds can be produced. Here we present a manufacturing process and culture conditions for fabrication of living machines using electrochemically compacted and aligned collagen (ELAC) as a scaffold. Using collagen as a scaffold results in a completely organic device. Milli-scale scaffolds were seeded with primary cardiomyocytes isolated from chick embryos and electrically stimulated to induce movement.

An essential component in the formation of understanding is the ability to use past experience to comprehend the here and now, and to aid selection of future action. Past experience is stored as memories which are then available for recall at very short notice, allowing for understanding of short and long term action. Autobiographical memory (ABM) is a form of temporally organised memory and is the organisation of episodes and contextual information from an individual’s experience into a coherent narrative, which is key to a sense of self. Formation and recall of memories is essential for effective and adaptive behaviour in the world, providing contextual information necessary for planning actions and memory functions, such as event reconstruction. Here we tested and developed a previously defined computational memory model, based on hippocampal structure and function, as a first step towards developing a synthetic model of human ABM (SAM). The hippocampal model chosen has functions analogous to that of human ABM. We trained the model on real-world sensory data and demonstrate successful, biologically plausible memory formation and recall, in a navigational task. The hippocampal model will later be extended for application in a biologically inspired system for human-robot interaction.

The cerebellum has an important role on motor learning. How sensory data arrives to the cerebellum is hardly understood. A two-phase model is proposed to understand how raw sensory data is processed to facilitate cerebellar predictive learning. Different candidates are presented for guiding the perceptual learning phase grounded on the role of the amygdala. A hebbian learning based computational model is presented with some preliminary results.

In this paper we present an architecture for the study of telepresence and human-robot interaction. The telepresence system uses the visual and gaze control systems of the iCub humanoid robot coupled with the Oculus Rift virtual reality system. The human is able to observe a remote location from the visual feedback displayed in the Oculus Rift. The exploration of the remote environment is achieved by controlling the eyes and head of the iCub humanoid robot with orientation information from human head movements. Our system was tested from various remote locations in a local network and through the internet, producing a smooth control of the robot. This provides a robust architecture for immersion of humans in a robotic system for remote observation and exploration of the environment.

In this paper, we present our work on the training of robotised architectural components of intelligent buildings, focusing on how architectural components can learn to behave animalistically, according to the judgment of human users. Our work aims at recovering the lost contact with animals in the urban context, taking advantage of biophilic empathy. The parameters governing the robotised elements we propose are mainly qualitative (emotions and aesthetical perception), which cannot easily be described by mathematical parameters. Additionally, due to their complexity, it is often impossible –or at least impractical, to hardcode suitable controllers for such structures. Thus, we propose the use of Artificial Intelligence learning techniques, concretely Evolutionary Algorithms, to allow the user to teach the robotised components how to behave in response to their resemblance to specific animal behaviors. This idea is tested on an intelligent façade that learns optimal configurations according to the perception of aggressiveness and calmness.