Next Generation Sensing Technology Lab

Kelley A10 | 4 East Alumni Avenue | Kingston, RI 02881

tao_wei@uri.edu – 401.874.2874

URI
Think Big, We Do.
Rhode Island Seal

Reflex-Tree

URI.Master

A novel sensing and control modality for future smart cities

 

Project Overview

  • Smart cites lie just on the horizon of 21st century urban life. Defined by their interconnected infrastructure and sensor systems, the technology market created by smart cites is projected to balloon to $1.5 trillion (Forbes, Frost and Sullivan) by the end of the decade.
  • This project studies a new computing and communication architecture, Reflex-Tree, with massive parallel sensing, data processing, and control functions designed to meet the challenges faced by future smart cities.
  • The central feature of this novel reflex-tree architecture is inspired by a fundamental element of the human nervous system — reflex arcs, or neuro-muscular reactions and instinctive motions in response to urgent situations that do not require the direct intervention of the brain.
  • The scientific foundation and engineering framework built by this project will pave the way for enhanced monitoring and management of critical smart city infrastructure, from gas/oil pipelines, water management, communication networks, and power grids, to public transportation and healthcare.Smart City Overview

 

Novel Reflex-Tree Architecture

reflect tree overview

The envisioned reflex-tree architecture, comprised of a 4-level hierarchy, is shown to the left. The lowest level, layer 4, contains a distributed sensor network with numerous sensory nodes that monitor public infrastructure by acquiring and processing critical data in parallel and in real time, allowing for unprecedented temporal and spatial resolution. Layer 3 consists of low-power, high-performance computing nodes, or edge devices. Each edge device is directly connected to a local sensor network (covering a neighborhood or a small community) and is responsible for raw data processing tasks, such as data classification and pattern recognition. Novel hardware architectures will be studied, allowing these edge devices to operate with maximal parallelism and pipelining with minimal power consumption. A cluster of such edge devices is then connected to an intermediate computing node that forms the next level of the hierarchy, layer 2. The key to this layer is our new spatial-temporal association based on the inputs from lower layers to support decision-making, and if necessary, to output such spatial-temporal patterns to the top layer for complex behavior analysis. Layer 1, the top level at the root of the reflex-tree, is the cloud with the high computing power necessary to provide citywide monitoring and control functions. Novel complex system behavior analysis and dynamic decision-making algorithms will be developed that are specifically tailored to heterogeneous accelerators. Efficient software runtime environments will be developed to optimize parallel processing, communication, and memory/storage hierarchy in the cloud. The result is a new computing platform with massive parallelism across all 4 layers, providing the necessary computing power and intelligence for future smart cities.

 

Research Methods

Igas linesn order to accomplish these ambitious yet feasible goals, this project will conduct fundamental scientific research and engineering development, leading to a special-purpose and massively parallel computing architecture. Central to this effort will be validation using a fully-functional, laboratory-scale municipal gas pipeline system. The system will include gate stations, distribution lines, regulators, and simulated loads and emergency events, all in an effort to develop increasingly effective methods of monitoring and controlling critical aspects of urban infrastructure.

 

Distributed Sensing Demonstration

Fiber optic sensing methods have several unique advantages that make them ideally suited for infrastructure monitoring. Chief among these advantages are high sensitivity, spacial resolution, and multiplexing capabilities, leading to the fabrication of thousands of sensitive detectors capable of localizing infrastructure changes in real time. The videos below demonstrate these important aspects, illustrating localized, real-time temperature sensing.

 

This video demonstrates the high spacial resolution of the sensing system, which is actively monitoring temperature along the fiber sensor.

 

This video demonstrates the physical mapping ability of the system, which is monitoring temperature over the length of the fiber sensor.

 

Current Progress

To date, our work has resulted in the publication of several scientific manuscripts thus far, with several others currently underway or actively under peer review. This work includes:

  • Bo Tang and Haibo He. “ENN: Extended Nearest Neighbor Method for Pattern Recognition [Research Frontier],” Computational Intelligence Magazine, IEEE, v.10, 2015, p. 52-60. doi:10.1109/MCI.2015.2437512
  • Chen, Z. and Hefferman, G. and Wei, T.. “Multiplexed oil level meter using a thin core fiber cladding mode exciter,” Photonics Technology Letters, IEEE, v.PP, 2015, p. 1-1. doi:10.1109/LPT.2015.2457295
  • Chen, Zhen and Hefferman, Gerald and Wei, Tao. “Multiplexed displacement fiber sensor using thin core fiber exciter,” Review of Scientific Instruments, v.86, 2015, p. -. doi:http://dx.doi.org/10.1063/1.4922019
  • Hefferman, G. and Zhen Chen and Lei Yuan and Tao Wei. “Phase-Shifted Terahertz Fiber Bragg Grating for Strain Sensing With Large Dynamic Range,” Photonics Technology Letters, IEEE, v.27, 2015, p. 1649-1652. doi:10.1109/LPT.2015.2433682
  • Hefferman, Gerald and Chen, Zhen and Wei, Tao. “Two-slot coiled coaxial cable resonator: Reaching critical coupling at a reduced number of coils,” Review of Scientific Instruments, v.85, 2014, p. -. doi:http://dx.doi.org/10.1063/1.4901593
  • Zhen Chen and Hefferman, G. and Lei Yuan and Yang Song and Tao Wei. “Ultraweak Waveguide Modification With Intact Buffer Coating Using Femtosecond Laser Pulses,” Photonics Technology Letters, IEEE, v.27, 2015, p. 1705-1708. doi:10.1109/LPT.2015.2438078
  • Zhen Chen and Lei Yuan and Gerald Hefferman and Tao Wei. “Ultraweak intrinsic Fabry-Perot cavity array for distributed sensing,” Opt. Lett., v.40, 2015, p. 320–323. doi:10.1364/OL.40.000320
  • Zhen Chen and Lei Yuan and Hefferman, G. and Tao Wei. “Terahertz Fiber Bragg Grating for Distributed Sensing,” Photonics Technology Letters, IEEE, v.27, 2015, p. 1084-1087. doi:10.1109/LPT.2015.2407580
  • Chen, Zhen and Rettinger, Ryan and Hefferman, Gerald and Smith, James and Oxley, Jimmie and Wei, Tao. “Microwave-Modulated Photon Doppler Velocimetry,” IEEE Photonics Technology Letters, v.28, 2016, p. 327–330.
  • Kay, Steven and Ding, Quan and Tang, Bo and He, Haibo. “Probability density function estimation using the EEF with application to subset/feature selection,” IEEE Transactions on Signal Processing, v.64, 2016, p. 641–651.
  • Tang, Bo and He, Haibo. “FSMJ: Feature Selection with Maximum Jensen-Shannon Divergence for Text Categorization,” arXiv preprint arXiv:1606.06366, 2016.
  • Tang, Bo and He, Haibo and Baggenstoss, Paul M and Kay, Steven. “A Bayesian classification approach using class-specific features for text categorization,” IEEE Transactions on Knowledge and Data Engineering, v.28, 2016, p. 1602–160.
  • Tang, Bo and Kay, Steven and He, Haibo. “Toward optimal feature selection in naive Bayes for text categorization,” IEEE Transactions on Knowledge and Data Engineering, v.28, 2016, p. 2508–252.
  • Tang, Bo and Kay, Steven and He, Haibo and Baggenstoss, Paul M. “EEF: Exponentially Embedded Families with Class-Specific Features for Classification,” IEEE Signal Processing Letters, v.23, 2016, p. 969–973.
  • Chen, Zhen and Hefferman, Gerald and Wei, Tao. “A low bandwidth DFB laser-based interrogator for terahertz-range fiber Bragg grating sensors,” IEEE Photonics Technology Letters, v.29, 2017, p. 365–368.
  • Chen, Zhen and Hefferman, Gerald and Wei, Tao. “Digitally controlled chirped pulse laser for sub-terahertz-range fiber structure interrogation,” Optics Letters, v.42, 2017, p. 1007–101.
  • Chen, Zhen and Hefferman, Gerald and Wei, Tao. “Field-programmable gate array-controlled sweep velocity-locked laser pulse generator,” Optical Engineering, v.56, 2017, p. 054102–0.
  • Chen, Zhen and Hefferman, Gerald and Wei, Tao. “Terahertz-range weak reflection fiber optic structures for sensing applications,” IEEE Journal of Selected Topics in Quantum Electronics, v.23, 2017, p. 1–6.
  • Hefferman, Gerald and Chen, Zhen and Wei, Tao. “Extended-bandwidth frequency sweeps of a distributed feedback laser using combined injection current and temperature modulation,” Review of Scientific Instruments, v.88, 2017, p. 075104.
  • Tang, Bo and Chen, Zhen and Hefferman, Gerald and Pei, Shuyi and Tao, Wei and He, Haibo and Yang, Qing. “Incorporating Intelligence in Fog Computing for Big Data Analysis in Smart Cities,” IEEE Transactions on Industrial Informatics, 2017.
  • D. Li, F. Wu, Y. Weng, Q. Yang and C. Xie. “HODS: Hardware Object Deserialization Inside SSD Storage,” 2018 IEEE 26th Annual International Symposium on Field-Programmable Custom Computing Machines (FCCM), 2018.
  • Liu, Tao and Wang, Feng and Zhang, Xuping and Yuan, Quan and Niu, Jihui and Zhang, Lin and Wei, Tao. “Interrogation of Ultra-Weak FBG Array Using Double-Pulse and Heterodyne Detection,” IEEE Photonics Technology Letters, v.30, 2018, p. 677–680.
  • X. Yang and H. He and X. Zhong. “Adaptive Dynamic Programming for Robust Regulation and Its Application to Power Systems,” IEEE Transactions on Industrial Electronics, v.65, 2018, p. 5722-5732. doi:10.1109/TIE.2017.2782205
  • Xiong Yang and Haibo He. “Self-learning robust optimal control for continuous-time nonlinear systems with mismatched disturbances,” Neural Networks, v.99, 2018, p. 19 – 30. doi:https://doi.org/10.1016/j.neunet.2017.11.022
  • J. Xu and B. Tang and H. He and H. Man. “Semisupervised Feature Selection Based on Relevance and Redundancy Criteria,” IEEE Transactions on Neural Networks and Learning Systems, v.PP, 2016, p. 1-11. doi:10.1109/TNNLS.2016.2562670
  • Tang, Bo and He, Haibo and Ding, Quan and Kay, Steven. “A Parametric Classification Rule Based on the Exponentially Embedded Family,” IEEE Transactions on Neural Networks and Learning Systems, v.26, 2015, p. 367–377.
  • Chen, Zhen and Hefferman, Gerald and Yuan, Lei and Song, Yang and Wei, Tao. “Terahertz-range interrogated grating-based two-axis optical fiber inclinometer,” Optical Engineering, v.55, 2016, p. 026106. doi:10.1117/1.OE.55.2.026106

 

WORKFORCE TRAINING

Brian (Chen) Zhen, PhD: System Engineer at Amazon

Bo Tang, PhD: Assistant Professor at University of Mississippi

Gerald Hefferman, MD : Transitional Resident Physician at Cambridge Health Alliance

Recent Comments

    Archives

    Categories

    • No categories

    Copyright © 2019 University of Rhode Island.

    The University of Rhode Island
    Think Big, We Do.
    A-ZDirectoryContact UsJump to top
    Close