Poster Session and Reception

Wednesday, September 6, 2017 - 3:30pm - 5:30pm
Lind 400
  • Sensor Location in Electrical Impedance Tomography Data collected during Pulmonary Function Tests
    Rashmi Murthy (Colorado State University)
    Electrical Impedance Tomography (EIT) is a low cost, portable and radiation free imaging modality that can be used to create low spatial resolution images based on varying electrical properties of biological tissues. An important application of EIT is pulmonary imaging. The images from EIT depend on the approximate knowledge of the position of the sensors. The boundary shape changes during the process of respiration and this in turn changes the knowledge of sensor location. In this poster we will present the EIT images that are obtained through D-bar algorithm, that incorporates the changes in the boundary shape during pulmonary function tests.
  • A Real-time Data Acquisition and Analytics System to Enhance Water Network Operation in Singapore
    Amin Rasekh (Sensus)
    Smart water networks are instrumented, interconnected water systems that integrate data, computation, control, and communication technologies. Smart water networks have ushered in a new era in design, operation, and management of urban water infrastructure. This poster provides an overview of the smart water network system performance in Singapore. This includes a networks of 210+ sensors that collect real-time, high-frequency data on water quality, pressure, flow and acoustic emission from the water supply network. This provides data for a wide range of real-time analytical capabilities to monitor, detect and notify on anomalies related to pressure variations, night flow, water quality issues, demand fluctuations and consumer-level Non-Revenue-Water tracking for revenue protection.
  • Experimental Design for Parameter Estimation in a System of Terrestrial Arthropods
    Amanda Laubmeier (North Carolina State University)
    There is an immediate need for theoretical models which can describe the effect of changing ecological communities, either due to species loss or migration, on trophic interactions. However, the experimental validation of such models is subject to practical limitations; ecological models might be opportunistically validated against data sets for which they were not intended, either due to cost or time constraints. With ecological collaborators at the Swedish University of Agricultural Sciences and California State University, Monterey Bay, we designed greenhouse experiments with the explicit intention of validating the Allometric Trophic Network (ATN) model, a Lotka-Volterra type model parameterized by body mass, for controlled insect ecosystems. We describe a PDE model for similar systems in landscape-dependent settings, with the eventual intention of designing a field experiment for further validation of predator-prey dynamics in a system of terrestrial arthropods.
  • Detecting methane super-emitters from the surface and space
    Daniel Cusworth (Harvard University)
    A large oil/gas field may include thousands of individual production wells along with gathering compressor stations, processing plants, and storage compressor stations. Methane flows through these different devices on its way to distribution pipelines. Malfunction of a device may cause anomalously high methane emissions, either intermittent or continuous. Recent field campaigns find that typically a few percent of devices in oil/gas fields have anomalously high methane emissions at any given time, and that these so-called “super-emitters” may account for over half of total emissions from oil/gas production. Satellites offer a particularly attractive resource for atmospheric monitoring because of their dense spatial coverage, but are limited by cloud cover. Surface sensor networks can also be effective for detection of super-emitters, in complement or independently from satellites. The number of sensors that can practically be deployed is typically small relative to the number of candidate super-emitters, so that the problem is underconstrained. We study how different configurations of continuous atmospheric observations from satellites and ground networks can serve to detect super-emitting devices in oil/gas fields. We use L-1 methods to pinpoint certain super-emitters in the inverse problem. For optimal sensor placement, we explore both distance-based clustering methods and placement based on the influence from a chemical transport model. Optimizing the network design of methane emitters provides an avenue optimize climate and financial benefits in the oil & gas sector.
  • Optimal Sensor Design for PDE Estimation
    Shuxia Tang (University of Waterloo)
    The development of smart materials makes it feasible to optimize sensor shapes. Optimization over shape is challenging though since the admissible set is infinite-dimensional. Kalman filters are optimal in the sense that they minimize the estimation error variance for given sensors. They are thus popular state estimators for both lumped and distributed parameter systems. Choosing the estimation error variance as the optimization criterion, conditions have been obtained that guarantee well-posedness of the optimal sensor design problem as well as continuity of the cost with respect to sensor design. A framework is constructed using finite-dimensional approximations for calculation of the optimal shape. Sufficient conditions are provided for the finite-dimensional optimal sensor configurations to converge to the infinite-dimensional optimal estimation performance.