Offloading and Transmission Strategies for IoT Edge Devices and Networks.

We present a machine and deep learning method to offload trained deep learning model and transmit packets efficiently on resource-constrained internet of things (IoT) edge devices and networks. Recently, the types of IoT devices have become diverse and the volume of data has been increasing, such as images, voice, and time-series sensory signals generated by various devices. However, transmitting large amounts of data to a server or cloud becomes expensive owing to limited bandwidth, and leads to latency for time-sensitive operations. Therefore, we propose a novel offloading and transmission policy considering energy-efficiency, execution time, and the number of generated packets for resource-constrained IoT edge devices that run a deep learning model and a reinforcement learning method to find an optimal contention window size for effective channel access using a contention-based medium access control (MAC) protocol. A Reinforcement learning is used to improve the performance of the applied MAC protocol. Our proposed method determines the offload and transmission strategies that are better to directly send fragmented packets of raw data or to send the extracted feature vector or the final output of deep learning networks, considering the operation performance and power consumption of the resource-constrained microprocessor, as well as the power consumption of the radio transceiver and latency for transmitting the all the generated packets. In the performance evaluation, we measured the performance parameters of ARM Cortex-M4 and Cortex-M7 processors for the network simulation. The evaluation results show that our proposed adaptive channel access and learning-based offload and transmission methods outperform conventional role-based channel access schemes. They transmit packets of raw data and are effective for IoT edge devices and network protocols.

Smartphone-Based Traveled Distance Estimation Using Individual Walking Patterns for Indoor Localization.

We introduce a novel method for indoor localization with the user's own smartphone by learning personalized walking patterns outdoors. Most smartphone and pedestrian dead reckoning (PDR)-based indoor localization studies have used an operation between step count and stride length to estimate the distance traveled via generalized formulas based on the manually designed features of the measured sensory signal. In contrast, we have applied a different approach to learn the velocity of the pedestrian by using a segmented signal frame with our proposed hybrid multiscale convolutional and recurrent neural network model, and we estimate the distance traveled by computing the velocity and the moved time. We measured the inertial sensor and global position service (GPS) position at a synchronized time while walking outdoors with a reliable GPS fix, and we assigned the velocity as a label obtained from the displacement between the current position and a prior position to the corresponding signal frame. Our proposed real-time and automatic dataset construction method dramatically reduces the cost and significantly increases the efficiency of constructing a dataset. Moreover, our proposed deep learning model can be naturally applied to all kinds of time-series sensory signal processing. The performance was evaluated on an Android application (app) that exported the trained model and parameters. Our proposed method achieved a distance error of <2.4% and >1.5% on indoor experiments.

An integrated microfluidic and fluorescence platform for probing in vivo neuropharmacology.

Neurotechnologies and genetic tools for dissecting neural circuit functions have advanced rapidly over the past decade, although the development of complementary pharmacological method-ologies has comparatively lagged. Understanding the precise pharmacological mechanisms of neuroactive compounds is critical for advancing basic neurobiology and neuropharmacology, as well as for developing more effective treatments for neurological and neuropsychiatric disorders. However, integrating modern tools for assessing neural activity in large-scale neural networks with spatially localized drug delivery remains a major challenge. Here, we present a dual microfluidic-photometry platform that enables simultaneous intracranial drug delivery with neural dynamics monitoring in the rodent brain. The integrated platform combines a wireless, battery-free, miniaturized fluidic microsystem with optical probes, allowing for spatially and temporally specific drug delivery while recording activity-dependent fluorescence using genetically encoded calcium indicators (GECIs), neurotransmitter sensors GRAB NE and GRAB DA , and neuropeptide sensors. We demonstrate the performance this platform for investigating neuropharmacological mechanisms in vivo and characterize its efficacy in probing precise mechanistic actions of neuroactive compounds across several rapidly evolving neuroscience domains.

Remote optogenetic control of the enteric nervous system and brain-gut axis in freely-behaving mice enabled by a wireless, battery-free optoelectronic device.

Wireless activation of the enteric nervous system (ENS) in freely moving animals with implantable optogenetic devices offers a unique and exciting opportunity to selectively control gastrointestinal (GI) transit in vivo, including the gut-brain axis. Programmed delivery of light to targeted locations in the GI-tract, however, poses many challenges not encountered within the central nervous system (CNS). We report here the development of a fully implantable, battery-free wireless device specifically designed for optogenetic control of the GI-tract, capable of generating sufficient light over large areas to robustly activate the ENS, potently inducing colonic motility ex vivo and increased propulsion in vivo. Use in in vivo studies reveals unique stimulation patterns that increase expulsion of colonic content, likely mediated in part by activation of an extrinsic brain-gut motor pathway, via pelvic nerves. This technology overcomes major limitations of conventional wireless optogenetic hardware designed for the CNS, providing targeted control of specific neurochemical classes of neurons in the ENS and brain-gut axis, for direct modulation of GI-transit and associated behaviours in freely moving animals.

Design of Center Pillar with Composite Reinforcements Using Hybrid Molding Method.

Recently, with the increase in awareness about a clean environment worldwide, fuel efficiency standards are being strengthened in accordance with exhaust gas regulations. In the automotive industry, various studies are ongoing on vehicle body weight reduction to improve fuel efficiency. This study aims to reduce vehicle weight by replacing the existing steel reinforcements in an automobile center pillar with a composite reinforcement. Composite materials are suitable for weight reduction because of their higher specific strength and stiffness compared to existing steel materials; however, one of the disadvantages is their high material cost. Therefore, a hybrid molding method that simultaneously performs compression and injection was proposed to reduce both process time and production cost. To replace existing steel reinforcements with composite materials, various reinforcement shapes were designed using a carbon fiber-reinforced plastic patch and glass fiber-reinforced plastic ribs. Structural analyses confirmed that, using these composite reinforcements, the same or a higher specific stiffness was achieved compared to the that of an existing center pillar using steel reinforcements. The composite reinforcements resulted in a 67.37% weight reduction compared to the steel reinforcements. In addition, a hybrid mold was designed and manufactured to implement the hybrid process.

A Study on the Mechanical Properties of an Automobile Part Additively Printed through Periodic Layer Rotation Strategies.

In metal product manufacturing, additive manufacturing (AM) is a method that has the advantage of fabricating complex shapes and customized production, unlike existing machining methods. However, owing to the characteristics of the AM process, anisotropy of macrostructure occurs because of various causes such as the scan direction, melting, fusion, and cooling of the powdered material. The macrostructure anisotropy is realized from the scan direction, and when a single layer is stacked in one direction, it is expressed as orthogonal anisotropy. Here, the classical lamination theory is applied to simply calculate the individual orthotropic layers by superimposing them. Through this, the authors analyzed whether the mechanical properties of the product are isotropically expressed with a periodic layer rotation strategy. To determine if the mechanical properties can be reasonably considered to be isotropic, a shock absorber mount for a vehicle was manufactured by AM. The tensile and vibration test performed on the product was compared with the finite element analysis and experimental results. As a result of the comparison, it was confirmed that the macroscopically of the product was considered isotropic as the load-displacement diagram and the fracture location coincided, as well as the natural frequency and mode shape.

Liquid-type Botulinum Toxin Type A in Adductor Spasmodic Dysphonia: A Prospective Pilot Study.

OBJECTIVES: Botulinum toxin (BTX) has been widely used to treat adductor spasmodic dysphonia (ADSD). Most commercially available forms of BTX require reconstitution before use, which may increase the risk of contamination and requires careful titration. Recently, a liquid-type BTX type A (BTX-A) has been developed, which should simplify the procedure and enhance its efficacy. Herein, we present a prospective pilot study to investigate the efficacy and safety of liquid-type BTX-A in the treatment of ADSD. METHODS: Twenty-six consecutive liquid-type BTX-A injections were performed in 12 patients with ADSD. We included as a control group 34 consecutive patients with ADSD who had previously undergone 52 vocal fold injection procedures with freeze-dried-type BTX-A. RESULTS: All patients in both groups had improvement of symptoms related to ADSD and period of normal voice. Most patients experienced breathiness, and the onset time, the peak response time, and the duration of breathiness were similar in both groups. The duration of effect (days) was 96.96 ± 18.91 and 77.38 ± 18.97 in the freeze-dried-type and the liquid-type groups, and the duration of benefit (days) was 80.02 ± 18.24 and 62.69 ± 19.73 in the freeze-dried-type and the liquid-type groups. To compare the efficacy between the freeze-dried-type and the liquid-type BTX-A, the sessions of the unilateral vocal fold injection were included and were categorized as group A (1 ~ 2 units BTX-A) and group B (2 ~ 3 units BTX-A), according to the dose per vocal fold. There was no significant difference of effect time between freeze-dried-type and liquid-type BTX-A groups. No adverse events related to BTX or vocal fold injection were reported. CONCLUSIONS: Liquid-type BTX-A is safe and effective for the treatment of spasmodic dysphonia. With the advantages of simple preparation, storage, and reuse and animal protein-free constituents, liquid-type BTX-A may be a good option in the treatment of spasmodic dysphonia.