A Scalable Neuro-inspired Robot Controller Integrating a Machine Learning Algorithm and a Spiking Cerebellar-Like Network

Ojeda, Ismael Baira; Tolu, Silvia; Lund, Henrik Hautop

doi:10.1007/978-3-319-63537-8_31

Cited by 4 publications

(3 citation statements)

References 26 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…While a global minimum is not guaranteed, but with an appropriate objective/fitness function, convergence towards a satisfying solution is achieved upon reaching a threshold value. The PSO acts to minimize the objective function, which is chosen as equation (17), to give good candidate solutions that minimize the mean error while reaching the targets. With a big enough population size and number of iterations/generations (relative to the number of parameters to be optimized in the search space), such an optimal solution/particle can be reached.…”

Section: Automated Tuning Of the Network Parametersmentioning

confidence: 99%

“…As in real biological systems, SNNs hold an advantage for real-time processing (as concluded in [13] for the visual system) and multiplexing of information (such as amplitude and frequency in the auditory system [14]). For robotic applications, the SNN allows building computational models of brain regions to imitate intelligent behavior in living organisms to a great extent, and in some cases even reveal the mysteries of the inner workings of the brain [15][16][17]. The currently rising neuromorphic chips allow real-time operation of such models while conserving power greatly compared to conventional systems [18].…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Differential mapping spiking neural network for sensor-based robot control

2021

View full text Add to dashboard Cite

In this work, a spiking neural network (SNN) is proposed for approximating differential sensorimotor maps of robotic systems. The computed model is used as a local Jacobian-like projection that relates changes in sensor space to changes in motor space. The SNN consists of an input (sensory) layer and an output (motor) layer connected through plastic synapses, with inter-inhibitory connections at the output layer. Spiking neurons are modeled as Izhikevich neurons with a synaptic learning rule based on spike timing-dependent plasticity. Feedback data from proprioceptive and exteroceptive sensors are encoded and fed into the input layer through a motor babbling process. A guideline for tuning the network parameters is proposed and applied along with the particle swarm optimization technique. Our proposed control architecture takes advantage of biologically plausible tools of an SNN to achieve the target reaching task while minimizing deviations from the desired path, and consequently minimizing the execution time. Thanks to the chosen architecture and optimization of the parameters, the number of neurons and the amount of data required for training are considerably low. The SNN is capable of handling noisy sensor readings to guide the robot movements in real-time. Experimental results are presented to validate the control methodology with a vision-guided robot.

show abstract

Section: Automated Tuning Of the Network Parametersmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Differential mapping spiking neural network for sensor-based robot control

2021

View full text Add to dashboard Cite

show abstract

“…The cerebellar SNN model included one plasticity site, at the cortical level, between PFs and PCs, based on a wellknown kind of STDP (38)(39)(40). Synaptic weights between PF-PC plasticity are modulated by IO activity (IOs-PCs connections in Table 2 are indicated as "teaching"), depending on the difference between the pre-and post-synaptic firing times (36,41,42). Long-Term Potentiation (LTP) and Long-Term Depression (LTD) are the two possible changes that each synaptic connection can undergo.…”

Section: Neurorobotic Implementationmentioning

confidence: 99%

Brain-inspired spiking neural network controller for a neurorobotic whisker system

Antonietti

Geminiani

Negri

et al. 2021

Preprint

View full text Add to dashboard Cite

It is common for animals to use self-generated movements to actively sense the surrounding environment. For instance, rodents rhythmically move their whiskers to explore the space close to their body. The mouse whisker system has become a standard model to study active sensing and sensorimotor integration through feedback loops. In this work, we developed a bioinspired spiking neural network model of the sensorimotor peripheral whisker system, modelling trigeminal ganglion, trigeminal nuclei, facial nuclei, and central pattern generator neuronal populations. This network was embedded in a virtual mouse robot, exploiting the Neurorobotics Platform, a simulation platform offering a virtual environment to develop and test robots driven by brain-inspired controllers. Eventually, the peripheral whisker system was properly connected to an adaptive cerebellar network controller. The whole system was able to drive active whisking with learning capability, matching neural correlates of behaviour experimentally recorded in mice.

show abstract

Combining Evolutionary and Adaptive Control Strategies for Quadruped Robotic Locomotion

et al. 2019

Self Cite

View full text Add to dashboard Cite

In traditional robotics, model-based controllers are usually needed in order to bring a robotic plant to the next desired state, but they present critical issues when the dimensionality of the control problem increases and disturbances from the external environment affect the system behavior, in particular during locomotion tasks. It is generally accepted that the motion control of quadruped animals is performed by neural circuits located in the spinal cord that act as a Central Pattern Generator and can generate appropriate locomotion patterns. This is thought to be the result of evolutionary processes that have optimized this network. On top of this, fine motor control is learned during the lifetime of the animal thanks to the plastic connections of the cerebellum that provide descending corrective inputs. This research aims at understanding and identifying the possible advantages of using learning during an evolution-inspired optimization for finding the best locomotion patterns in a robotic locomotion task. Accordingly, we propose a comparative study between two bio-inspired control architectures for quadruped legged robots where learning takes place either during the evolutionary search or only after that. The evolutionary process is carried out in a simulated environment, on a quadruped legged robot. To verify the possibility of overcoming the reality gap, the performance of both systems has been analyzed by changing the robot dynamics and its interaction with the external environment. Results show better performance metrics for the robotic agent whose locomotion method has been discovered by applying the adaptive module during the evolutionary exploration for the locomotion trajectories. Even when the motion dynamics and the interaction with the environment is altered, the locomotion patterns found on the learning robotic system are more stable, both in the joint and in the task space.

show abstract

A Scalable Neuro-inspired Robot Controller Integrating a Machine Learning Algorithm and a Spiking Cerebellar-Like Network

Cited by 4 publications

References 26 publications

Differential mapping spiking neural network for sensor-based robot control

Differential mapping spiking neural network for sensor-based robot control

Brain-inspired spiking neural network controller for a neurorobotic whisker system

Combining Evolutionary and Adaptive Control Strategies for Quadruped Robotic Locomotion

Contact Info

Product

Resources

About