Jumping and landing movements are characterized by large instantaneous forces, short duration, and a high uncertainty concerning take off and landing points. Such characteristics make conventional types of control and robot design inadequate. Here we present an approach to realize motor control of jumping and landing which exploits the synergy between control and mechanical structure. Our experimental system is a pneumatically actuated bipedal robot called "Mowgli". Mowgli's artificial musculoskeletal system consists of six McKibben pneumatic muscle actuators including bi-articular muscle and two legs with hip, knee, and ankle joints. Mowgli can reach jump heights of more than 50% of its body height and can land softly. Our results show a proximo-distal sequence of joint extensions during jumping despite simultaneous motor activity. Extensions in the whole body motion are caused by the compliance and the natural dynamics of the legs. In addition to the experiments with the real robot, we also simulated two types of open loop controllers for vertical jumping with disturbance. We found that the model controlled by open loop motor command through a muscle-tendon mechanism could jump robustly. The simulation results demonstrate the contribution of the artificial musculoskeletal system as a physical feedback loop in explosive movements.

Cortical learning via sensorimotor experiences evoked by bodily movements begins as early as the foetal period. However, the learning mechanisms by which sensorimotor experiences guide cortical learning remain unknown owing to technical and ethical difficulties. To bridge this gap, we present an embodied brain model of a human foetus as a coupled brain-body-environment system by integrating anatomical/physiological data. Using this model, we show how intrauterine sensorimotor experiences related to bodily movements induce specific statistical regularities in somatosensory feedback that facilitate cortical learning of body representations and subsequent visual-somatosensory integration. We also show how extrauterine sensorimotor experiences affect these processes. Our embodied brain model can provide a novel computational approach to the mechanistic understanding of cortical learning based on sensorimotor experiences mediated by complex interactions between the body, environment and nervous system.

Replacing sensors with indirect sensing techniques contributes to retaining the flexibility of soft robots. By combining physical models with recurrent neural networks (which we term a physics-informed recurrent neural network [PIRNN] approach), we implemented a hybrid prediction scheme on two typical soft pneumatic actuators: a McKibben pneumatic artificial muscle and a pneumatic-based soft finger made of silicone. The results showed that this hybrid scheme robustly enhanced the prediction accuracy to a great extent, even when combined with an inaccurate physical model. We also present the broad applicability of the PIRNN approach, showing its effectiveness for diverse types of RNNs and soft robotics platforms. Our work fills the gaps in the literature by applying a physics-informed machine-learning approach to practical engineering problems in soft robotics.

Long-horizon dexterous robot manipulation of deformable objects, such as banana peeling, is a problematic task because of the difficulties in object modeling and a lack of knowledge about stable and dexterous manipulation skills. This paper presents a goal-conditioned dual-action (GC-DA) deep imitation learning (DIL) approach that can learn dexterous manipulation skills using human demonstration data. Previous DIL methods map the current sensory input and reactive action, which often fails because of compounding errors in imitation learning caused by the recurrent computation of actions. The method predicts reactive action only when the precise manipulation of the target object is required (local action) and generates the entire trajectory when precise manipulation is not required (global action). This dual-action formulation effectively prevents compounding error in the imitation learning using the trajectory-based global action while responding to unexpected changes in the target object during the reactive local action. The proposed method was tested in a real dual-arm robot and successfully accomplished the banana-peeling task.

An approach based on reservoir computing enables design of chaotic itinerancy reproducibly in a high-dimensional chaotic system.