MSGID: 1:317/3 64a6437b   
   PID: hpt/lnx 1.9.0-cur 2019-01-08   
   TID: hpt/lnx 1.9.0-cur 2019-01-08   
    Growing bio-inspired polymer brains for artificial neural networks   
      
      
     Date:   
         July 5, 2023   
     Source:   
         Osaka University   
     Summary:   
         A new method for connecting neurons in neuromorphic wetware has been   
         developed. The wetware comprises conductive polymer wires grown in   
         a three-dimensional configuration, done by applying square-wave   
         voltage to electrodes submerged in a precursor solution. The   
         voltage can modify wire conductance, allowing the network to be   
         trained. This fabricated network is able to perform unsupervised   
         Hebbian learning and spike-based learning.   
      
      
         Facebook Twitter Pinterest LinkedIN Email   
      
   ==========================================================================   
   FULL STORY   
   ==========================================================================   
   A new method for connecting neurons in neuromorphic wetware has   
   been developed by researchers from Osaka University and Hokkaido   
   University. The wetware comprises conductive polymer wires grown in a   
   three-dimensional configuration, done by applying square-wave voltage to   
   electrodes submerged in a precursor solution. The voltage can modify wire   
   conductance, allowing the network to be trained. This fabricated network   
   is able to perform unsupervised Hebbian learning and spike-based learning.   
      
   The development of neural networks to create artificial intelligence in   
   computers was originally inspired by how biological systems work. These   
   'neuromorphic' networks, however, run on hardware that looks nothing   
   like a biological brain, which limits performance. Now, researchers from   
   Osaka University and Hokkaido University plan to change this by creating   
   neuromorphic 'wetware'.   
      
   While neural-network models have achieved remarkable success in   
   applications such as image generation and cancer diagnosis, they still   
   lag far behind the general processing abilities of the human brain. In   
   part, this is because they are implemented in software using traditional   
   computer hardware that is not optimized for the millions of parameters   
   and connections that these models typically require.   
      
   Neuromorphic wetware, based on memristive devices, could address this   
   problem.   
      
   A memristive device is a device whose resistance is set by its history   
   of applied voltage and current. In this approach, electropolymerization   
   is used to link electrodes immersed in a precursor solution using wires   
   made of conductive polymer. The resistance of each wire is then tuned   
   using small voltage pulses, resulting in a memristive device.   
      
   "The potential to create fast and energy-efficient networks has been shown   
   using 1D or 2D structures," says senior author Megumi Akai-Kasaya. "Our   
   aim was to extend this approach to the construction of a 3D network."   
   The researchers were able to grow polymer wires from a common polymer   
   mixture called 'PEDOT:PSS', which is highly conductive, transparent,   
   flexible, and stable. A 3D structure of top and bottom electrodes was   
   first immersed in a precursor solution. The PEDOT:PSS wires were then   
   grown between selected electrodes by applying a square-wave voltage   
   on these electrodes, mimicking the formation of synaptic connections   
   through axon guidance in an immature brain.   
      
   Once the wire was formed, the characteristics of the wire, especially   
   the conductance, were controlled using small voltage pulses applied   
   to one electrode, which changes the electrical properties of the film   
   surrounding the wires.   
      
   "The process is continuous and reversible," explains lead author Naruki   
   Hagiwara, "and this characteristic is what enables the network to be   
   trained, just like software-based neural networks."  The fabricated   
   network was used to demonstrate unsupervised Hebbian learning (i.e.,   
   when synapses that often fire together strengthen their shared connection   
   over time). What's more, the researchers were able to precisely control   
   the conductance values of the wires so that the network could complete   
   its tasks. Spike-based learning, another approach to neural networks that   
   more closely mimics the processes of biological neural networks, was also   
   demonstrated by controlling the diameter and conductivity of the wires.   
      
   Next, by fabricating a chip with a larger number of electrodes and   
   using microfluidic channels to supply the precursor solution to each   
   electrode, the researchers hope to build a larger and more powerful   
   network. Overall, the approach determined in this study is a big step   
   toward the realization of neuromorphic wetware and closing the gap   
   between the cognitive abilities of humans and computers.   
      
       * RELATED_TOPICS   
             o Matter_&_Energy   
                   # Electronics # Telecommunications # Fuel_Cells #   
                   Spintronics   
             o Computers_&_Math   
                   # Neural_Interfaces # Computers_and_Internet #   
                   Computer_Programming # Spintronics_Research   
       * RELATED_TERMS   
             o Neural_network o Artificial_neural_network o Mobile_phone o   
             Transformer o Local_area_network o Voice_over_IP o Computer_worm   
             o Nanowire   
      
   ==========================================================================   
      
    Print   
      
    Email   
      
    Share   
   ==========================================================================   
   ****** 1 ****** ***** 2 ***** **** 3 ****   
   *** 4 *** ** 5 ** Breaking this hour   
   ==========================================================================   
       * Why_Birds_Ancestors_Lived;_Other_Dinosaurs_Died *   
       Dissolving_Cardiac_Device_Treats_Heart_Disease *   
       Webb_Locates_Dust_Reservoirs_in_Two_Supernovae *   
       Earth_Formed_from_Dry,_Rocky_Building_Blocks *   
       Ancient_Volcanic_Activity_On_Moon's_Dark_Side *   
       Highly_Conductive_Metallic_Gel_for_3D_Printing *   
       Potent_Greenhouse_Gas_Could_Be_Abated_Today *   
       Polymer_Brains_for_Artificial_Neural_Networks *   
       Early_Apex_Predator_Sought_Soft_Over_...   
      
       * Time_in_Universe_Once_Flowed_Five_Times_Slower   
      
   Trending Topics this week   
   ==========================================================================   
   SPACE_&_TIME Black_Holes Astrophysics NASA MATTER_&_ENERGY Biochemistry   
   Optics Petroleum COMPUTERS_&_MATH Communications Educational_Technology   
   Computer_Modeling   
      
      
   ==========================================================================   
      
   Strange & Offbeat   
   ==========================================================================   
   SPACE_&_TIME   
   Quasar_'Clocks'_Show_Universe_Was_Five_Times_Slower_Soon_After_the_Big_Bang   
   First_'Ghost_Particle'_Image_of_Milky_Way   
   Gullies_on_Mars_Could_Have_Been_Formed_by_Recent_Periods_of_Liquid_Meltwater,   
   Study_Suggests MATTER_&_ENERGY   
   Researchers_Create_Highly_Conductive_Metallic_Gel_for_3D_Printing   
   Growing_Bio-Inspired_Polymer_Brains_for_Artificial_Neural_Networks   
   Displays_Controlled_by_Flexible_Fins_and_Liquid_Droplets_More_Versatile,   
   Efficient_Than_LED_Screens COMPUTERS_&_MATH   
   AI_Tests_Into_Top_1%_for_Original_Creative_Thinking   
   Turning_Old_Maps_Into_3D_Digital_Models_of_Lost_Neighborhoods   
   NeuWS_Camera_Answers_'Holy_Grail_Problem'_in_Optical_Imaging Story Source:   
   Materials provided by Osaka_University. Note: Content may be edited for   
   style and length.   
      
      
   ==========================================================================   
   Journal Reference:   
      1. Naruki Hagiwara, Tetsuya Asai, Kota Ando, Megumi Akai‐Kasaya.   
      
         Fabrication and Training of 3D Conductive Polymer Networks for   
         Neuromorphic Wetware. Advanced Functional Materials, 2023; DOI:   
         10.1002/ adfm.202300903   
   ==========================================================================   
      
   Link to news story:   
   https://www.sciencedaily.com/releases/2023/07/230705105850.htm   
      
   --- up 1 year, 18 weeks, 2 days, 10 hours, 50 minutes   
    * Origin: -=> Castle Rock BBS <=- Now Husky HPT Powered! (1:317/3)   
   SEEN-BY: 15/0 106/201 114/705 123/120 153/7715 218/700 226/30 227/114   
   SEEN-BY: 229/110 112 113 307 317 400 426 428 470 664 700 291/111 292/854   
   SEEN-BY: 298/25 305/3 317/3 320/219 396/45 5075/35   
   PATH: 317/3 229/426