Design and Implementation of a Modular Biomimetic Infochemical Communication System

Abstract

We describe here the design and implementation of a novel biomimetic infochemical communication system that employs airborne molecules alone to communicate over space and time. The system involves the design and fabrication of a microsystem capable of producing and releasing a precise mix of biosynthetic compounds and a sensor system capable of detecting and decoding the ratiometrically encoded chemical information.The research inspired by biology has been based upon the biosynthetic pathways of infochemical production and information processing within the insect world. In this novel approach, the functional equivalents of the nanoscale biological machinery are implemented by combining the latest advances and convergence of expertise in the fields of biochemistry, molecular biology, neuroscience, micro-and nanofabrication, materials science, and smart sensor and microcircuit design. The biomimetic system comprises a micromachined bio-reactor mimicking the sex gland of the female insect that releases a blend of pheromones in precisely controlled ratios, together with a cell-based biosensor system, mimicking the antennae of the male insect. The signals from the biosensors are classified and ratios decoded using a field-programmable gate array implementation of a neuromorphic model of the antenna lobe of the insect.We believe that this novel, smart infochemical communication system, inspired by the insect's behavior, could eventually be implemented in VLSI technology at low cost and low power with possible application in the fields of automatic identification and data capturing, product labeling, search and rescue, environmental monitoring, and pest control.

Rapid processing of chemosensor transients in a neuromorphic implementation of the insect macroglomerular complex

“…Exploiting LCA dynamics was found to improve the time where the classification rose above chance, producing a more rapid classification in ratio, allowing also a more accurate classification with P classification for the MGC model outputs above that of the normalized chemosensor data across all time points. Such temporally sensitive neuromorphic processing shown by our MGC AL model could become extremely valuable for improved chemical sensing through stimulus-dependent spatiotemporal patterning imposed on the chemosensor array to enhance discrimination (Gardner et al, 2007; Sánchez-Montañés et al, 2008), and also as part of an infochemical communication system in which rapid and robust transient processing is paramount (Cole et al, 2009a,b; Rácz et al, 2011, 2013). …”

“…(C) Schematic of the neuromorphic MGC model showing the programmable logic neuronal model components and their interconnectivity—triangles correspond to leaky integrate-and-fire (LIF) neuron models, circles to exponential decay synapses and elliptical “+” elements correspond to current synaptic integration nodes. (D) Field programmable gate array (FPGA) floorplan of the physical device showing the footprint for the MGC model optimized down to 3095 of the 11,200 available logic slices of a Xilinx Virtex 5 device [ B and D reprinted with permission from Rácz et al (2013)].…”

“…The polymer coating thicknesses were estimated based on the measured frequency changes (145 kHz for PCL and 111 kHz for PVK) to be only 15 nm for PCL and 11 nm for PVK using the calculation method described in (Sang-Hun et al, 2005). Further details on the SAWR chemical sensors are provided in (Rácz et al, 2013; Yang et al, 2012). …”

“…Using ratios in this way provides a robust encoding mechanism when communicating that we have exploited in the first example of an infochemical communication system based upon both moth biosynthesis and ratiometric chemoreception (Cole et al, 2009a,b; Rácz et al, 2011, 2013). This technological objective together with understanding the underlying neural mechanisms for ratiometric decoding, motivate this study as the insect MGC system provides an ideal model system to investigate the close relationships between external input stimuli and internal brain dynamics.…”

High frequency surface acoustic wave resonator-based sensor for particulate matter detection

“…If the devices are placed in an oscillator configuration these changes can be measured with great accuracy as a change in resonant frequency. The sensing mechanism of the surface wave sensor depends on various factors including mechanical and electrical properties of the particles under test [14,15]. The mechanical effects include mass density and visco-elasticity of the particles, while the electrical effects include their conductivity and dielectric permittivity [16].…”