Physical Substrates Research Articles

Modelling and evaluating restricted ESNs on single- and multi-timescale problems

AbstractReservoir Computing is a computing model ideal for performing computation on varied physical substrates. However, these physical reservoirs can be difficult to scale up. We propose joining various reservoirs together as an approach to solving this problem, simulating physical reservoirs with Echo State Networks (ESNs). We investigate various methods of combining ESNs to form larger reservoirs, including a method that we dub Restricted ESNs. We provide a notation for describing Restricted ESNs, and use it to benchmark a standard ESN against restricted ones. We investigate two methods to keep the weight matrix density consistent when comparing a Restricted ESN to a standard one, which we call overall consistency and patch consistency. We benchmark restricted ESNs on NARMA10 and the sunspot prediction benchmark, and find that restricted ESNs perform similarly to standard ones. We present some application scenarios in which restricted ESNs may offer advantages over standard ESNs. We then test restricted ESNs on a version of the multi-timescale Multiple Superimposed Sines tasks, in order to establish a baseline performance that can be improved upon in further work. We conclude that we can scale up reservoir performance by linking small homogeneous subreservoirs together without significant loss in performance over a single large reservoir, justifying future work on using heterogeneous subreservoirs for greater flexibility.

Quaternarized chitosan nanofiber and ZIF aerogel composites for synergetic CO2 cycloaddition catalysis

Chemical upcycling of CO2, a major greenhouse gas, is attracting significant attention as a crucial strategy to combat global warming. The production of cyclic carbonate using metal-organic frameworks and their composites using nanofibrous carbohydrate polymer are promising ways to convert CO2 into valuable products. However, the current role of fibrous polymers is restricted to serving as physical substrates. This study seeks to expand the functionality of quaternarized chitosan nanofibers into synergistic catalysts in addition to their physical support role. A novel aerogel composite using Co/Zn-ZIF catalyst and quaternarized chitosan nanofiber (Q-CsNF+) was fabricated, and its hierarchical pore structure was extensively discussed. The obtained ZIF/Q-CsNF+ composite can synergistically convert epoxide to cyclic carbonate by acting as a co-catalyst. Moreover, we determined the predominant factors influencing catalytic activity in CO2 cycloaddition, especially by examining the interplay between CO2 affinity and co-catalyst effects. This research provides fundamental insight into developing CO2 cycloaddition catalysts using nature-derived fibrous polymers, opening new avenues for sustainable and efficient CO2 utilization.

Morphological, mechanical and biological assessment of PCL/pristine graphene scaffolds for bone regeneration

Scaffolds are physical substrates for cell attachment, proliferation, and differentiation, ultimately leading to the regeneration of tissues. They must be designed according to specific biomechanical requirements such as mechanical properties, surface characteristics, biodegradability, biocompatibility, and porosity. The optimal design of a scaffold for a specific tissue strongly depends on both materials and manufacturing processes. Polymeric scaffolds reinforced with electro-active particles could play a key role in tissue engineering by modulating cell proliferation and differentiation. This paper investigates the use of an extrusion additive manufacturing system to produce PCL/pristine graphene scaffolds for bone tissue applications. PCL/pristine graphene blends were prepared using a melt blending process. Scaffolds with regular and reproducible architecture were produced with different concentrations of pristine graphene. Scaffolds were evaluated from morphological, mechanical, and biological view. The results suggest that the addition of pristine graphene improves the mechanical performance of the scaffolds, reduces the hydrophobicity, and improves cell viability and proliferation.

Exploring the Architectural Biases of the Canonical Cortical Microcircuit.

The cortex plays a crucial role in various perceptual and cognitive functions, driven by its basic unit, the canonical cortical microcircuit. Yet, we remain short of a framework that definitively explains the structure-function relationships of this fundamental neuroanatomical motif. To better understand how physical substrates of cortical circuitry facilitate their neuronal dynamics, we employ a computational approach using recurrent neural networks and representational analyses. We examine the differences manifested by the inclusion and exclusion of biologically-motivated inter-areal laminar connections on the computational roles of different neuronal populations in the microcircuit of two hierarchically-related areas, throughout learning. Our findings show that the presence of feedback connections correlates with the functional modularization of cortical populations in different layers, and provides the microcircuit with a natural inductive bias to differentiate expected and unexpected inputs at initialization. Furthermore, when testing the effects of training the microcircuit and its variants with a predictive-coding inspired strategy, we find that doing so helps better encode noisy stimuli in areas of the cortex that receive feedback, all of which combine to suggest evidence for a predictive-coding mechanism serving as an intrinsic operative logic in the cortex.

Deep Photonic Reservoir Computer for Speech Recognition.

Speech recognition is a critical task in the field of artificial intelligence (AI) and has witnessed remarkable advancements thanks to large and complex neural networks, whose training process typically requires massive amounts of labeled data and computationally intensive operations. An alternative paradigm, reservoir computing (RC), is energy efficient and is well adapted to implementation in physical substrates, but exhibits limitations in performance when compared with more resource-intensive machine learning algorithms. In this work, we address this challenge by investigating different architectures of interconnected reservoirs, all falling under the umbrella of deep RC (DRC). We propose a photonic-based deep reservoir computer and evaluate its effectiveness on different speech recognition tasks. We show specific design choices that aim to simplify the practical implementation of a reservoir computer while simultaneously achieving high-speed processing of high-dimensional audio signals. Overall, with the present work, we hope to help the advancement of low-power and high-performance neuromorphic hardware.

Criticality in FitzHugh-Nagumo oscillator ensembles: Design, robustness, and spatial invariance

Reservoir computing is an efficient and flexible framework for decision-making, control, and signal processing. It uses a network of interacting components varying from abstract nonlinear dynamical systems to physical substrates. Despite recent progress, the hardware implementation with inherent parameter variability and uncertainties, such as those mimicking the properties of living organisms’ nervous systems, remains an active research area. To address these challenges, we propose a constructive approach using a network of FitzHugh-Nagumo oscillators, exhibiting criticality across a broad range of resistive coupling strengths and robustness without specific parameter tuning. Additionally, the network’s activity demonstrates spatial invariance, offering freedom in choosing readout nodes. We introduce an alternative characterization of criticality by analyzing power dissipation, and demonstrate that criticality supports the robustness of the classification accuracy with respect to the readout shrinkage. Our results indicate criticality as a valuable property for classification problems, and provides design concepts for bio-inspired computational paradigms.

Harnessing synthetic active particles for physical reservoir computing

The processing of information is an indispensable property of living systems realized by networks of active processes with enormous complexity. They have inspired many variants of modern machine learning, one of them being reservoir computing, in which stimulating a network of nodes with fading memory enables computations and complex predictions. Reservoirs are implemented on computer hardware, but also on unconventional physical substrates such as mechanical oscillators, spins, or bacteria often summarized as physical reservoir computing. Here we demonstrate physical reservoir computing with a synthetic active microparticle system that self-organizes from an active and passive component into inherently noisy nonlinear dynamical units. The self-organization and dynamical response of the unit are the results of a delayed propulsion of the microswimmer to a passive target. A reservoir of such units with a self-coupling via the delayed response can perform predictive tasks despite the strong noise resulting from the Brownian motion of the microswimmers. To achieve efficient noise suppression, we introduce a special architecture that uses historical reservoir states for output. Our results pave the way for the study of information processing in synthetic self-organized active particle systems.

In‐Sensor Passive Speech Classification with Phononic Metamaterials

AbstractMitigating the energy requirements of artificial intelligence requires novel physical substrates for computation. Phononic metamaterials have vanishingly low power dissipation and hence are a prime candidate for green, always‐on computers. However, their use in machine learning applications has not been explored due to the complexity of their design process. Current phononic metamaterials are restricted to simple geometries (e.g., periodic and tapered) and hence do not possess sufficient expressivity to encode machine learning tasks. A non‐periodic phononic metamaterial, directly from data samples, that can distinguish between pairs of spoken words in the presence of a simple readout nonlinearity is designed and fabricated, hence demonstrating that phononic metamaterials are a viable avenue towards zero‐power smart devices.

Instant Detection of Synthetic Cannabinoids on Physical Matrices, Implemented on a Low-Cost, Ultraportable Device.

Synthetic cannabinoids (SCs) make up a class of novel psychoactive substances (NPS), used predominantly in prisons and homeless communities in the U.K. SCs can have severe side effects, including psychosis, stroke, and seizures, with numerous reported deaths associated with their use. The chemical diversity of SCs presents the major challenge to their detection since approaches relying on specific molecular recognition become outdated almost immediately. Ideally one would have a generic approach to detecting SCs in portable settings. The problem of SC detection is more challenging still because the majority of SCs enter the prison estate adsorbed onto physical matrices such as paper, fabric, or herb materials. That is, regardless of the detection modality used, the necessary extraction step reduces the effectiveness and ability to rapidly screen materials on-site. Herein, we demonstrate a truly instant generic test for SCs, tested against real-world drug seizures. The test is based on two advances. First, we identify a spectrally silent region in the emission spectrum of most physical matrices. Second, the finding that background signals (including from autofluorescence) can be accurately predicted is based on tracking the fraction of absorbed light from the irradiation source. Finally, we demonstrate that the intrinsic fluorescence of a large range of physical substrates can be leveraged to track the presence of other drugs of interest, including the most recent iterations of benzodiazepines and opioids. We demonstrate the implementation of our presumptive test in a portable, pocket-sized device that will find immediate utility in prisons and law enforcement agencies around the world.

State-of-the-art and prospective progress of growing AlN substrates by physical vapor transport

SiC, GaN, diamond, and AlN known as wide band-gap semiconductors, integrate the advantages of microelectronics, optoelectronics, and power electronics. The ultra-wide bandgap semiconductor AlN crystal has a direct bandgap of 6.1 eV at room temperature. AlN has led to a burgeoning research interest in the fabrication of devices due to its outstanding structure and fascinating physiochemical (microelectronics, optoelectronics and power-electronics, etc.) properties. In this review, we summarize some recent advances in growing physical vapor transport AlN substrates. First, we introduce the primary seed growth of AlN using two disparate growth strategies: on heterogeneous substrate and spontaneous nucleation. We then elaborate on their advantages and disadvantages for subsequent homogeneous growth. Second, we highlight ultramodern advances that have been made in the preparation of high-quality and large-size AlN substrates, discuss the optical properties of AlN. In the end, the current challenges and new opportunities for PVT grown AlN substrates in improving the quality are also discussed.

A Brief History of Transcranial Magnetic Stimulation

A Brief History of Transcranial Magnetic Stimulation

Joint QoS and energy-efficient resource allocation and scheduling in 5G Network Slicing

Network Slicing (NS) is fast evolving as a prominent enabler for providing tailored services in the Fifth Generation of cellular networks (5G). Network Slices are virtualized network entities formed over physical substrates, deployed for the customized application use cases. A Network Slice needs to exhibit end to end capabilities and meet Quality of Service (QoS) specifications and Service Level Agreements (SLAs). To provide end-to-end traffic management capabilities in the network slice, firstly, traffic flows are categorized into different priority traffic classes, and their severity levels are assessed. Priorities can be applied across cellular and IP based systems. Machine Learning (ML) algorithms are employed on QoS profile attributes in establishing traffic priorities in slices. Secondly, we propose a novel algorithm for NS Resource Partitioning and User Allocation. We put forward an online virtual backbone based solution for resource allocation and priority class-based packet scheduling. This joint QoS and energy efficiency driven approach is built on top of established traffic classes and dynamic power savings techniques. Finally, through Cognitive Cycles (CC), we devise better network re-configuration to obtain more energy savings. Traffic classifier modules are implemented using Jupyter notebook and Python API. Scheduling and resource allocation modules in networks slices are emulated in Mininet, Flowvisor, and Beacon and POX controllers. The simulation results reveal the reduced node consumption is achieved through the evolutionary CC algorithm, and it outperforms other standard approaches by at least 23%. Similarly, for the traffic priority prediction, from the results, we could infer Gradient Boosting and Random Forest Regressors exhibit superior accuracy with the root mean square deviation of 2.2% and 1.2% respectively when compared to other standard ML algorithms.

Energy Efficiency and Delay Optimization of Virtual Slicing of Fog Radio Access Network

To develop the energy efficient of 6G networks, the fog radio access network (F-RAN) is expected to meet various use cases of high-performance mobile services. Although network virtualization greatly enhances the flexibility to accommodate various services, the interaction and optimization between delay and energy efficiency (EE) in virtual slicing (VS)-based F-RAN have not been well studied. To accomplish the EE and delay optimization of VS of F-RAN. The key technical challenges lie in the construction of new network architecture, the integration and optimization of radio, caching, and computing 3-D resources, reducing the algorithm’s complexity, and simulation verification. We first design a novel network architecture based on VS and fog computing. The VS method in F-RAN to embed VS assembles virtual radio, caching, and computing resources into physical substrates, transforming nonconvex problems into convex optimization problems by transforming constraints and using Lyapunov optimization methods. We further propose a virtual resource allocation optimization algorithm based on EE. To reduce the complexity of the algorithm, a low-complexity EE optimization algorithm is further proposed for virtual resource allocation. The simulation results show that the low-complexity virtual resource allocation EE optimization algorithm proposed in this article has better performance than the existing fog access network resource allocation methods. The EE is further improved by about 30% under the condition that the guaranteed delay threshold is two slots (i.e., 1 ms).

Integrated world modeling theory expanded: Implications for the future of consciousness.

Integrated world modeling theory (IWMT) is a synthetic theory of consciousness that uses the free energy principle and active inference (FEP-AI) framework to combine insights from integrated information theory (IIT) and global neuronal workspace theory (GNWT). Here, I first review philosophical principles and neural systems contributing to IWMT's integrative perspective. I then go on to describe predictive processing models of brains and their connections to machine learning architectures, with particular emphasis on autoencoders (perceptual and active inference), turbo-codes (establishment of shared latent spaces for multi-modal integration and inferential synergy), and graph neural networks (spatial and somatic modeling and control). Future directions for IIT and GNWT are considered by exploring ways in which modules and workspaces may be evaluated as both complexes of integrated information and arenas for iterated Bayesian model selection. Based on these considerations, I suggest novel ways in which integrated information might be estimated using concepts from probabilistic graphical models, flow networks, and game theory. Mechanistic and computational principles are also considered with respect to the ongoing debate between IIT and GNWT regarding the physical substrates of different kinds of conscious and unconscious phenomena. I further explore how these ideas might relate to the "Bayesian blur problem," or how it is that a seemingly discrete experience can be generated from probabilistic modeling, with some consideration of analogies from quantum mechanics as potentially revealing different varieties of inferential dynamics. I go on to describe potential means of addressing critiques of causal structure theories based on network unfolding, and the seeming absurdity of conscious expander graphs (without cybernetic symbol grounding). Finally, I discuss future directions for work centered on attentional selection and the evolutionary origins of consciousness as facilitated "unlimited associative learning." While not quite solving the Hard problem, this article expands on IWMT as a unifying model of consciousness and the potential future evolution of minds.

Measuring and monitoring restored ecosystems: can remote sensing be applied to the ecological recovery wheel to inform restoration success?

The commencement of the United Nations Decade on Ecosystem Restoration has highlighted the urgent need to improve restoration science and fast‐track ecological outcomes. The application of remote sensing for monitoring purposes has increased over the past two decades providing a variety of image datasets and derived products suitable to map and measure ecosystem properties (e.g. vegetation species, community composition, and structural dimensions such as height and cover). However, the operational use of remote sensing data and derived products for ecosystem restoration monitoring in research, industry, and government has been relatively limited and underutilized. In this paper, we use the Society for Ecological Restoration (SER) ecological recovery wheel (ERW) to assess the current capacity of drone‐airborne‐satellite remote sensing datasets to measure each of the SER's recommended attributes and sub‐attributes for terrestrial restoration projects. Based on our combined expertise in the areas of ecological monitoring and remote sensing, a total of 11 out of 18 sub‐attributes received the highest feasibility score and show strong potential for remote sensing assessments; while sub‐attributes such as gene flows, all trophic levels and chemical and physical substrates have a reduced capacity for monitoring. We argue that in the coming decade, ecologists can combine remote sensing with the ERW to monitor restoration recovery and reference ecosystems for improved restoration outcomes at the local, regional, and landscape scales. The ERW approach can be adapted as a monitoring framework for projects to utilize the benefits of remote sensing and inform management through scalable, operational, and meaningful outcomes.

Guest Editorial Special Issue on New Frontiers in Extremely Efficient Reservoir Computing

With the penetration of artificial intelligence (AI) technology into industrial applications, not only computational effectiveness but also computational efficiency in machine learning (ML) methods has been increasingly demanded. Reservoir computing (RC) is an ML framework leveraging a dynamic <i>reservoir</i> for a nonlinear transformation of sequential inputs and a <i>readout</i> for mapping the reservoir state to a desired output. Since only the readout is trained with a simple learning algorithm, RC has attracted much attention as a promising approach to enhance compatibility between high computational performance and low learning cost. In addition, recent studies on physical reservoirs implemented with various physical substrates have boosted the potential of RC in the development of effective and efficient AI hardware. Therefore, it is time to further explore the new frontiers in extremely efficient RC.

Manipulative repertoire of bonobos (Pan paniscus) in spontaneous feeding situation.

Comparative behavioral studies of hand use amongst primate species, including humans, have been central in research on evolutionary mechanisms. In particular, the manipulative abilities of our closest relatives, the chimpanzee (Pan troglodytes), have been widely described in various contexts, showing a high level of dexterity both in zoo and in natural conditions. In contrast, the study of bonobos' manipulative abilities has almost exclusively been carried out in experimental contexts related to tool use. The objective of the present study is to describe the richness of the manipulative repertoire of zoo-housed bonobos, in a spontaneous feeding context including various physical substrates to gain a larger insight into our evolutionary past. Our study describes a great variety of grasping postures and grip associations in bonobos, close to the range of manipulative repertoire in chimpanzees, confirming that the two species are not markedly different in terms of cognitive and morphological constraints associated with food manipulation. We also observed differences in manipulative behaviors between juveniles and adults, indicating a greater diversity in grip associations and grasping postures used in isolation with age, and a sex-biased use of tools with females using tools more often than males. These results are consistent with the previous results in the Pan genus and reinforce the hypothesis that the evolutionary mechanisms underlying the flexibility of manipulative behaviors are shared by both species and that these ecological strategies would have already evolved in their common ancestor.

Initial wood trait variation overwhelms endophyte community effects for explaining decay trajectories

AbstractMicrobial organisms, environmental conditions and their interactions govern many ecosystem processes. Recent studies have highlighted the importance of priority effects, that is, the identity of potential decomposers present early in community assembly, in determining resulting decay rates especially for wood. In diverse forests, available woody substrates differ chemically and structurally with implications for their role as both habitats and resources for microbes. Both wood traits and microbial communities at the start of the decay process affect subsequent decay rates, but the relative magnitude of effects is not known. In this work, we sought to ask a simple question: what are the relative effects of microbial communities and wood traits?We characterized fungal and oomycete endophytes with amplicon sequencing from stems of 22 woody species growing in woodlands near Richmond, NSW, Australia, and measured 11 traits to capture variation in the physical and chemical wood substrates. To evaluate the consequence of endophyte diversity and wood traits on the trajectory of decay, stem samples were sequentially harvested over 5 years to quantify the decay rate, its consistency and how it varies through time.We did not find evidence to support particular initial endophyte compositions leading to faster decay. Instead, initial wood attributes were much more powerful in explaining decay trajectories with smaller, less dense stems with high water, low N and low lignin concentrations decomposing consistently faster.These data show that initial wood traits have long‐lasting consequences on decay unlike natural variation in endophyte communities, supporting the idea that community member functions are highly redundant. Wood substrate‐driven environmental filtering, rather than endophyte‐driven priority effects, had a stronger effect on decay when real‐world levels of diversity in wood traits were considered.Read the freePlain Language Summaryfor this article on the Journal blog.

Deep physical neural networks trained with backpropagation

Deep-learning models have become pervasive tools in science and engineering. However, their energy requirements now increasingly limit their scalability1. Deep-learning accelerators2–9 aim to perform deep learning energy-efficiently, usually targeting the inference phase and often by exploiting physical substrates beyond conventional electronics. Approaches so far10–22 have been unable to apply the backpropagation algorithm to train unconventional novel hardware in situ. The advantages of backpropagation have made it the de facto training method for large-scale neural networks, so this deficiency constitutes a major impediment. Here we introduce a hybrid in situ–in silico algorithm, called physics-aware training, that applies backpropagation to train controllable physical systems. Just as deep learning realizes computations with deep neural networks made from layers of mathematical functions, our approach allows us to train deep physical neural networks made from layers of controllable physical systems, even when the physical layers lack any mathematical isomorphism to conventional artificial neural network layers. To demonstrate the universality of our approach, we train diverse physical neural networks based on optics, mechanics and electronics to experimentally perform audio and image classification tasks. Physics-aware training combines the scalability of backpropagation with the automatic mitigation of imperfections and noise achievable with in situ algorithms. Physical neural networks have the potential to perform machine learning faster and more energy-efficiently than conventional electronic processors and, more broadly, can endow physical systems with automatically designed physical functionalities, for example, for robotics23–26, materials27–29 and smart sensors30–32.

"Hitchhicking with invertebrates": two reports of epibiosis by peritrich ciliates on ostracods and hydrachnid mites in tanks of epiphytic bromeliads from south Brazil.

Temporary waters are common environments found in physical and biological substrates. Among them, some bromeliads species are known to hold water in their tanks, in a habitat called phytotelmata. Phytotelmata serve as habitats for several organisms, from bacteria and protists to arthropods and anurans. Peritrich ciliates are often found as epibionts on aquatic invertebrates in these environments. Here, we report two cases of epibiosis involving Lagenophrys sp. attached to ostracods (Elpidium spp.) and Rhabdostyla sp. colonizing hydrachnid mites in the tanks of two bromeliad species. In our analysis, we measured the frequency of epibiosis considering the presence of both basibiont and epibiont in the samples. The results shown a significant difference between Elpidium sp. and Lagenophrys sp. compared to hydrachnid mites and Rhabdostyla sp. (87.5% and 19%, respectively), supported by the Kruskal-Walis test (p = 0.0003, Chi-square = 9.687). These reports are important since the knowledge of phytotelmata communities from tropical and subtropical areas is incipient, although it has been increasing over the last years. It also shows that epibiosis doesn't always represent a beneficial relationship. These two epibiosis systems found in bromeliad tanks raise questions about organism's dispersal throughout other phytotelmata and other temporary water habitats.