Lidé

Ing. Rudolf Jakub Szadkowski

Všechny publikace

Bootstrapping the Dynamic Gait Controller of the Soft Robot Arm

  • Autoři: Ing. Rudolf Jakub Szadkowski, Nazeer, M.S., Cianchetti, M., Falotico, E., prof. Ing. Jan Faigl, Ph.D.,
  • Publikace: 2023 IEEE International Conference on Robotics and Automation. Piscataway: IEEE, 2023. p. 2669-2675. ISSN 2577-087X. ISBN 979-8-3503-2365-8.
  • Rok: 2023
  • DOI: 10.1109/ICRA48891.2023.10160579
  • Odkaz: https://doi.org/10.1109/ICRA48891.2023.10160579
  • Pracoviště: Katedra počítačů, Centrum umělé inteligence
  • Anotace:
    In this paper, we propose a novel dynamic gait controller for the repetitive behavior of soft robot manipulators performing routine tasks. Compliance with soft robots is advantageous when the robot interacts with living organisms and other fragile objects. However, predicting and controlling repetitive behavior is challenging because of hysteresis and non-linear dynamics governing the interactions. Existing priorfree methods track the dynamic state using recurrent neural networks or rely on known generalized coordinates describing the robot's state. We propose to model the interaction induced by the repetitive behavior as gait dynamics and represent the dynamic state with Central Pattern Generator (CPG) tracking the motion phase and thus reduce the complexity of the robot's forward model. The proposed method bootstraps an ensemble of the forward models exploring multiple dynamic contexts that are expanded as it searches for repetitive motion producing the target repetitive behavior. The proposed approach is experimentally validated on a pneumatically actuated soft robot arm I-Support, where the method infers gaits for different targets.

Hexapod Gait Control Through Internal Model Belief Update

  • DOI: 10.18910/92290
  • Odkaz: https://doi.org/10.18910/92290
  • Pracoviště: Katedra počítačů, Centrum umělé inteligence
  • Anotace:
    Gait is a common locomotion strategy for legged ani-mals where a repetitive sequence of leg movements results in motion. Assuming the predictive coding theory, the agent continually represents its beliefs about the gait state in a state estimation [1]. In Internal Model (IM) principle, the state is predicted from efference copy by a Forward Model (FM) that can be further utilized for motor control [2]. However, in gait dynamics, the sensory state depends on the sequence of the previous motor commands, which makes the sensory- motor dynamics challenging to model.

Continually trained life-long classification

  • DOI: 10.1007/s00521-021-06154-9
  • Odkaz: https://doi.org/10.1007/s00521-021-06154-9
  • Pracoviště: Centrum umělé inteligence
  • Anotace:
    Two challenges can be found in a life-long classifier that learns continually: the concept drift, when the probability distribution of data is changing in time, and catastrophic forgetting when the earlier learned knowledge is lost. There are many proposed solutions to each challenge, but very little research is done to solve both challenges simultaneously. We show that both the concept drift and catastrophic forgetting are closely related to our proposed description of the life-long continual classification. We describe the process of continual learning as a wrap modification, where a wrap is a manifold that can be trained to cover or uncover a given set of samples. The notion of wraps and their cover/uncover modifiers are theoretical building blocks of a novel general life-long learning scheme, implemented as an ensemble of variational autoencoders. The proposed algorithm is examined on evaluation scenarios for continual learning and compared to state-of-the-art algorithms demonstrating the robustness to catastrophic forgetting and adaptability to concept drift but also showing the new challenges of the life-long classification.

Deep Transfer Learning of Traversability Assessment for Heterogeneous Robots

  • Pracoviště: Centrum umělé inteligence
  • Anotace:
    For autonomous robots operating in an unknown environment, it is important to assess the traversability of the surrounding terrain to improve path planning and decision-making on where to navigate next in a cost-efficient way. Specifically, in mobile robot exploration, terrains and their traversability are unknown prior to the deployment. The robot needs to use its limited resources to learn its terrain traversability model on the go; however, reusing a provided model is still a desirable option. In a team of heterogeneous robots, the models assessing traversability cannot be reused directly since robots might possess different morphology or sensory equipment and thus experience the terrain differently. In this paper, we propose a transfer learning approach for convolutional neural networks assessing the traversability between heterogeneous robots, where the transferred network is retrained using data available for the target robot to accommodate itself to the robot’s traversability. The proposed method is verified in real-world experiments, where the proposed approach provides faster learning convergence and better traversal cost predictions than the baseline.

Gait Adaptation After Leg Amputation of Hexapod Walking Robot Without Sensory Feedback

  • DOI: 10.1007/978-3-031-15934-3_54
  • Odkaz: https://doi.org/10.1007/978-3-031-15934-3_54
  • Pracoviště: Centrum umělé inteligence
  • Anotace:
    In this paper, we address the adaptation of the locomotion controller to change of the multi-legged walking robot morphology, such as leg amputation. In nature, the animal compensates for the amputation using its neural locomotion controller that we aim to reproduce with the Central Pattern Generator (CPG). The CPG is a rhythm-generating recurrent neural network used in gait controllers for the rhythmical locomotion of walking robots. The locomotion corresponds to the robot's morphology, and therefore, the locomotion rhythm must adapt if the robot's morphology is changed. The leg amputation can be handled by sensory feedback to compensate for the load distribution imbalances. However, the sensory feedback can be disrupted due to unexpected external events causing the leg to be damaged, thus leading to unexpected motion states. Therefore, we propose dynamic rules for learning a new gait rhythm without the sensory feedback input. The method has been experimentally validated on a real hexapod walking robot to demonstrate its usability for gait adaptation after amputation of one or two legs.

Learning-based Detection of Leg-Surface Contact using Position Feedback Only

  • DOI: 10.1109/ETFA52439.2022.9921720
  • Odkaz: https://doi.org/10.1109/ETFA52439.2022.9921720
  • Pracoviště: Centrum umělé inteligence
  • Anotace:
    In this work-in-progress report, we present experimental results of lightweight learning-based leg-contact detection methods for a small hexapod walking robot with position feed- back only. The detection of the leg contact with the surface is addressed as anomaly detection using predicted and measured positions of the leg’s joints in the leg swing phase. A polynomial regressor and three-layer neural network are evaluated regarding the prediction error and computational requirements using realistic datasets collected with the real hexapod walking robot.

Terrain Traversal Cost Learning with Knowledge Transfer Between Multi-legged Walking Robot Gaits

  • DOI: 10.1109/ICARSC55462.2022.9784790
  • Odkaz: https://doi.org/10.1109/ICARSC55462.2022.9784790
  • Pracoviště: Centrum umělé inteligence
  • Anotace:
    The terrain traversal abilities of multi-legged walking robots are affected by gaits, the walking patterns that enable adaptation to various operational environments. Fast and lowset gaits are suited to flat ground, while cautious and highset gaits enable traversing rough areas. A suitable gait can be selected using prior experience with a particular terrain type. However, experience alone is insufficient in practical setups, where the robot experiences each terrain with only one or just a few gaits and thus would infer novel gait-terrain interactions from insufficient data. Therefore, we use knowledge transfer to address unsampled gait-terrain interactions and infer the traversal cost for every gait. The proposed solution combines gaitterrain cost models using inferred gait-to-gait models projecting the robot experiences between different gaits. We implement the cost models as Gaussian Mixture regressors providing certainty to identify unknown terrains where knowledge transfer is desirable. The presented method has been verified in synthetic showcase scenarios and deployment with a real walking robot. The proposed knowledge transfer demonstrates improved cost prediction and selection of the appropriate gait for specific terrains.

Traversability Transfer Learning Between Robots with Different Cost Assessment Policies

  • DOI: 10.1007/978-3-030-98260-7_21
  • Odkaz: https://doi.org/10.1007/978-3-030-98260-7_21
  • Pracoviště: Centrum umělé inteligence
  • Anotace:
    Predicting mobile robots' traversability over terrains is crucial to select safe and efficient paths through rough and unstructured environments. In multi-robot missions, knowledge transfer techniques can enable learning terrain traversability assessment the robots did not experience individually. The knowledge can be incrementally aggregated for homogeneous robots since they can treat foreign knowledge as their own. However, robots with different perceptions might experience the same terrain differently, so it is impossible to aggregate the shared knowledge directly. In this paper, we show how to learn a model that transfers the experience between heterogeneous robots, enabling each robot to use the whole sum of the experience of the multi-robot team. The proposed approach uses correlation to combine individual neural networks that assess the traversability of individual robots. The presented method has been verified in a real-world deployment of multi-legged walking robots with different cost assessment policies.

Experimental Leg Inverse Dynamics Learning of Multi-legged Walking Robot

  • DOI: 10.1007/978-3-030-70740-8_10
  • Odkaz: https://doi.org/10.1007/978-3-030-70740-8_10
  • Pracoviště: Katedra počítačů, Centrum umělé inteligence
  • Anotace:
    Rough terrain locomotion is a domain where multi-legged robots benefit from their relatively complex morphology compared to the wheeled or tracked robots. Efficient rough terrain locomotion requires the legged robot sense contacts with the terrain to adapt its behavior and cope with the terrain irregularities. Usage of inverse dynamics to estimate the leg state and detect the leg contacts with the terrain suffers from computational complexity. Furthermore, it requires a precise analytical model identification that does not cope with adverse changes of the leg parameters such as friction changes due to the joint wear, the increased weight of the leg due to the mud deposits, and possible leg morphology change due to damage. In this paper, we report the experimental study on the locomotion performance with machine learning-based inverse dynamics model learning. Experimental examining three different learning models show that a simplified model is sufficient for leg collision detection learning. Moreover, the learned model is faster for calculation and generalizes better than more complex models when the leg parameters change.

Gait Genesis Through Emergent Ordering of RBF Neurons on Central Pattern Generator for Hexapod Walking Robot

  • Pracoviště: Katedra počítačů, Centrum umělé inteligence
  • Anotace:
    The neurally based gait controllers for multi-legged robots are designed to reproduce the plasticity observed in animal locomotion. In animals, gaits are regulated by Central Pattern Generator (CPG), a recurrent neural network producing rhythmical signals prescribing each leg’s action timing, leading to coordinated motion of multiple legs. The biomimetic CPG-RBF architecture, where leg motion timing is encoded by Radial Basis Function (RBF) neurons coupled with CPG, is used in recent gait controllers. However, the RBF neurons coupling is usually parameterized by the supervisor. Therefore, the RBF parameters get outdated when the CPG signal’s wave-form changes. We propose self-supervised dynamics for RBF parameters adapting to a given CPG and producing the required gait rhythm. The method orders the leg activity with respect to inter-leg coordination rules and maps the activity onto CPG states. The proposed dynamics produce rhythmic control for three different hexapod gaits and adapts to the CPG parametric changes.

Self-Learning Event Mistiming Detector Based on Central Pattern Generator

  • DOI: 10.3389/fnbot.2021.629652
  • Odkaz: https://doi.org/10.3389/fnbot.2021.629652
  • Pracoviště: Katedra počítačů, Centrum umělé inteligence
  • Anotace:
    A repetitive movement pattern of many animals, a gait, is controlled by the Central Pattern Generator (CPG), providing rhythmic control synchronous to the sensed environment. As a rhythmic signal generator, the CPG can control the motion phase of biomimetic legged robots without feedback. The CPG can also act in sensory synchronization, where it can be utilized as a sensory phase estimator. Direct use of the CPG as the estimator is not common, and there is little research done on its utilization in the phase estimation. Generally, the sensory estimation augments the sensory feedback information, and motion irregularities can reveal from comparing measurements with the estimation. In this work, we study the CPG in the context of phase irregularity detection, where the timing of sensory events is disturbed. We propose a novel self-supervised method for learning mistiming detection, where the neural detector is trained by dynamic Hebbian-like rules during the robot walking. The proposed detector is composed of three neural components: (i) the CPG providing phase estimation, (ii) Radial Basis Function neuron anticipating the sensory event, and (iii) Leaky Integrate-and-Fire neuron detecting the sensory mistiming. The detector is integrated with the CPG-based gait controller. The mistiming detection triggers two reflexes: the elevator reflex, which avoids an obstacle, and the search reflex, which grasps a missing foothold. The proposed controller is deployed and trained on a hexapod walking robot to demonstrate the mistiming detection in real locomotion. The trained system has been examined in the controlled laboratory experiment and real field deployment in the Bull Rock cave system, where the robot utilized mistiming detection to negotiate the unstructured and slippery subterranean environment.

Neurodynamic Sensory-Motor Phase Binding for Multi-Legged Walking Robots

  • DOI: 10.1109/IJCNN48605.2020.9207507
  • Odkaz: https://doi.org/10.1109/IJCNN48605.2020.9207507
  • Pracoviště: Katedra počítačů, Centrum umělé inteligence
  • Anotace:
    Motivated by observations of animal behavior, locomotion of multi-legged walking robots can be controlled by the central pattern generators (CPGs) that produce a repetitive motion pattern. A rhythmic pattern, a gait, is defined by phase relations between all leg joints. In a case of an external influence such as terrain irregularity, some actuator phase can shift and thus disrupt the phase relations between the actuators. The actuator phase relations can be maintained only by synchronizing to the sensors, which output can indicate the motion disruption. However, establishing correct sensory-motor phase relations requires not only the motor phase model but also a model of the sensory phase, which is generally unknown. Although both sensory and motor phases can be modeled by single CPG, the capabilities of such CPG-based controllers are limited because they are not flexible and robust. In this paper, we propose to model the phases of each sensor and motor by separate CPGs. The phase relations between the sensor and motor phases are established by radial basis function (RBF) neurons learned with proposed periodic Grossberg rule for which we present the convergence proof. Based on the reported evaluation results using high-fidelity simulation, the proposed locomotion controller demonstrates the desired plasticity, and it is capable of learning multiple gaits with robust synchronization to terrain changes using sensor inputs.

Transfer of Inter-Robotic Inductive Classifier

  • DOI: 10.1109/ICACR51161.2020.9265509
  • Odkaz: https://doi.org/10.1109/ICACR51161.2020.9265509
  • Pracoviště: Katedra počítačů, Centrum umělé inteligence
  • Anotace:
    In multi-robot deployments, the robots need to share and integrate their own experience and perform transfer learning. Under the assumption that the robots have the same morphology and carry equivalent sensory equipment, the problem of transfer learning can be considered incremental learning. Thus, the transfer learning problem inherits the challenges of incremental learning, such as catastrophic forgetting and concept drift. In catastrophic forgetting, the model abruptly forgets the previously learned knowledge during the learning process. The concept drift arises with different experiences between consecutively sampled models. However, state-of-the-art robotic transfer learning approaches do not address both challenges at once. In this paper, we propose to use an incremental classifier on a transfer learning problem. The feasibility of the proposed approach is demonstrated in a real deployment. The robot consistently merges two classifiers learned on two different tasks into a classifier that performs well on both tasks.

Autoencoders Covering Space as a Life-Long Classifier

  • DOI: 10.1007/978-3-030-19642-4_27
  • Odkaz: https://doi.org/10.1007/978-3-030-19642-4_27
  • Pracoviště: Katedra počítačů, Centrum umělé inteligence
  • Anotace:
    A life-long classifier that learns incrementally has many challenges such as concept drift, when the class changes in time, and catastrophic forgetting when the earlier learned knowledge is lost. Many successful connectionist solutions are based on an idea that new data are learned only in a part of a network that is relevant to the new data. We leverage this idea and propose a novel method for learning an ensemble of specialized autoencoders. We interpret autoencoders as manifolds that can be trained to contain or exclude given points from the input space. This manifold manipulation allows us to implement a classifier that can suppress catastrophic forgetting and adapt to concept drift. The proposed algorithm is evaluated on an incremental version of the XOR problem and on an incremental version of the MNIST classification where we achieved 0.9 accuracy which is a significant improvement over the previously published results

Basic Evaluation Scenarios for Incrementally Trained Classifiers

  • DOI: 10.1007/978-3-030-30484-3_41
  • Odkaz: https://doi.org/10.1007/978-3-030-30484-3_41
  • Pracoviště: Centrum umělé inteligence
  • Anotace:
    Evaluation of incremental classification algorithms is a complex task because there are many aspects to evaluate. Besides the aspects such as accuracy and generalization that are usually evaluated in the context of classification, we also need to assess how the algorithm handles two main challenges of the incremental learning: the concept drift and the catastrophic forgetting. However, only catastrophic forgetting is evaluated by the current methodology, where the classifier is evaluated in two scenarios for class addition and expansion. We generalize the methodology by proposing two new scenarios of incremental learning for class inclusion and separation that evaluate the handling of the concept drift. We demonstrate the proposed methodology on the evaluation of three different incremental classifiers, where we show that the proposed methodology provides a more complete and finer evaluation.

Learning Central Pattern Generator Network with Back-Propagation Algorithm

  • Autoři: Ing. Rudolf Jakub Szadkowski, Čížek, P., prof. Ing. Jan Faigl, Ph.D.,
  • Publikace: Proceedings of the 18th Conference Information Technologies - Applications and Theory (ITAT 2018). Aachen: CEUR Workshop Proceedings, 2018. p. 116-123. vol. 2203. ISSN 1613-0073. ISBN 9781727267198.
  • Rok: 2018
  • Pracoviště: Katedra počítačů, Centrum umělé inteligence
  • Anotace:
    An adaptable central pattern generator (CPG) that di-rectly controls the rhythmic motion of multilegged robotmust combine plasticity and sustainable periodicity. Thiscombination requires an algorithm that searches the para-metric space of the CPG and yields a non-stationary andnon-divergent solution. We model the CPG with the pi-oneering Matsuoka’s neural oscillator which is (mostly)non-divergent and provides constraints ensuring non-stationarity. We embed these constraints into the CPGformulation which we further implemented as a layer ofan artificial neural network. This enables the CPG to belearnable by back-propagation algorithm while sustainingthe desirable properties. Moreover, the proposed CPG canbe integrated into more complex networks and trained un-der different optimization objectives. In addition to thetheoretical properties of the developed system, its flexibil-ity is demonstrated in successful learning of the tripod mo-tion gait with its practical deployment on the real hexapodwalking robot.

Terrain Classification with Crawling Robot Using Long Short-Term Memory Network

  • DOI: 10.1007/978-3-030-01424-7_75
  • Odkaz: https://doi.org/10.1007/978-3-030-01424-7_75
  • Pracoviště: Katedra počítačů, Centrum umělé inteligence
  • Anotace:
    Terrain classification is a crucial feature for mobile robots operating across multiple terrains. One way to learn a terrain classifier is to use a stream of labeled proprioceptive data recorded during a terrain traversal. In this paper, we propose a new terrain classifier that combines a feature extraction from a data stream with the long short-term memory (LSTM) network. Features are extracted from the information-sparse data stream by applying a sliding window computing three central moments. The feature sequence is continuously classified by the LSTM network into multiple terrain classes. Furthermore, a modified bagging method is used to deal with a limited and unbalanced training set. In comparison to the previous work on terrain classifiers for a hexapod crawling robot using only servo-drive feedback, the proposed classifier provides continuous classification with the F1 score up to 0.88, and thus provide better results than SVM classifier learned on the same input data.

Za stránku zodpovídá: Ing. Mgr. Radovan Suk