Ted's Home Page Get Firefox! Get Thunderbird! Friday, October 24, 2014, 12:41 pm GMT
    Home :: Contact :: Blog :: Updates (Thu, Sep 29) :: Search :: Highlight Linux :: Apache HTTPD :: PHP :: VIM :: OCS
          email me email me
 
[Ted and a baby giraffe]
 
Contact Me
Résumé/CV
More Information
 
Faculty Job Search
Industry Job Search
Publication Directory
Research/Graduate
Work
My Teaching Sites
 
General Posts
(including
LaTeX
templates)
 
Utilities
 
Public hg repos
Public git repos
 
Links
The Blog
 
Guestbook
 
 
 
 
   
 
 
  email me
email me
email me
email me

Ted :: Early Research (Graduate School)

NOTICE: I AM LOOKING FOR A RESEARCH/TEACHING JOB! (tenure-track assistant professor)

At the moment, I am seeking an engineering (electrical or computer science) or applied mathematics tenure-track assistant professor position that will let me continue scholarly research in distributed systems. I have expertise in biomimicry, optimal control, hybrid systems, parallel algorithms, and agent-based modeling. My application areas span intelligent lighting, robotics, and autonomous vehicles. I also have extensive teaching expertise in electrical engineering, including circuits, signals and systems, and control theory courses.

A presentation of my faculty application materials is available at
http://www.tedpavlic.com/jobsearch/
This site may be a better place to start for some readers.

Research Overview, Observations, and Goals
  • Short overview of completed research and research goals

  • "Engineering Serendipity: Successes in Solitary Foraging and New Investigations in Cooperative Task-Processing Networks" (LARGE PDF FILE: 4.5 MB) – Slides from a research seminar surveying biomimicry results and observations about biomimicry in general
    • Abstract:
      Biomimicry has been a great source of novel ideas for bad technological design. Both engineers and biologists look for concrete examples where behavioral models can be applied directly. Consequently, engineers overlook important theoretical assumptions behind those models, and biologists overlook non-traditional design problems because of the attraction to applications with robots, the charismatic megafauna of technology. Rather than using concrete applications to drive interdisciplinary research, abstract mathematical models can be generated to fit a range of disciplines. Valuable insights into both design and analysis then come from studying the assumptions necessary to specialize these models. Additionally, advances in one field can better translate to another when there are frameworks and communication shared between disciplines.

      First in this talk, the consequences of generalizing a simple model of solitary foraging from behavioral ecology is discussed. In particular, the process produces decision-making algorithms for autonomous search as well as new explanations for both impulsiveness and overstaying observed in animals. The talk then introduces a generic model of asynchronous distributed cooperative task processing that is meant to mimic human, non-human, and artificial systems. Each agent is interested in maximizing its own utility; however, patterns qualitatively similar to load balancing occur. Although the model has little resemblance to the birth--death processes typically described in network cooperation theory, a version of the networked generalization of Hamilton's rule emerges.
    • LaTeX source code (Mercurial repository; powerdot-based presentation)



Foraging Theory and Engineering

Early Proposed Plan of Research (PDF) (HTML)

Master's Thesis (Optimal Foraging Theory Revisited)
Related Books (not authored by me)


Coperative Task Processing

Doctoral Candidacy Exam
  • PhD research proposal: "Cooperative Task Processing" (PDF) (source code repository)
  • Written candidacy exam submission: "Research Problems in Distributed Control for Energy Systems" (PDF) (source code repository)

  • Internal research group presentation (given later in January of 2010) of research results: "Cooperative Task Processing: A Framework" (PDF) (source code repository – Powerdot source code)


Design and Analysis of Optimal Task-Processing Agents

Doctoral Dissertation
  • Dissertation: Design and Analysis of Optimal Task-Processing Agents (PDF [LARGE 2 MB]) (OhioLink) (source code repository)
    • Abstract:
      This dissertation is given in two parts followed by concluding remarks. The first three chapters describe the generalization of optimal foraging theory for the design of solitary task-processing agents. The following two chapters address the coordinated action of distributed independent agents to achieve a desirable global result. The short concluding part summarizes contributions and future research directions.

      Optimal foraging theory (OFT) uses ecological models of energy intake to predict behaviors favored by natural selection. Using models of the long-term rate of energetic gain of a solitary forager encountering a variety of food opportunities at a regular rate, it predicts characteristics of optimal solutions that should be expressed in nature. Several engineered agents can be modeled similarly. For example, an autonomous air vehicle (AAV) that flies over a region encounters targets randomly just as an animal will encounter food as it travels. OFT describes the preferences that the animal is likely to have due to natural selection. Thus, OFT applied to mobile vehicles describes the preferences of successful vehicle designs.

      Although OFT has had success in existing engineering applications, rate maximization is not a good fit for many applications that are otherwise analogous to foraging. Thus, in the first part of this dissertation, the classical OFT methods are rediscovered for generic optimization objectives. It is shown that algorithms that are computationally equivalent to those inspired by classical OFT can perform better in realistic scenarios because they are based on more feasible optimization objectives. It is then shown how the design of foraging-like algorithms provides new insight into behaviors observed and expected in animals. The generalization of the classical methods extracts fundamental properties that may have been overlooked in the biological case. Consequently, observed behaviors that have been previously been called irrational are shown to follow from the extension of the classical methods.

      The second part of the dissertation describes individual agent behaviors that collectively result in the achievement of a global optimum when the distributed agents operate in parallel. In the first chapter, collections of agents that are each similar to the agents from the early chapters are considered. These agents have overlapping capabilities, and so one agent can share the task processing burden of another. For example, an AAV patrolling one area can request the help of other vehicles patrolling other areas that have a sparser distribution of targets. We present a method of volunteering to answer the request of neighboring agents such that sensitivity to the relative loading across the network emerges. In particular, agents that are relatively more loaded answer fewer task-processing requests and receive more answers to their own requests. The second chapter describes a distributed numerical optimization method for optimization under inseparable constraints. Inseparable constraints typically require some direct coordination between distributed solver agents. However, we show how certain implementations allow for stigmergy, and so far less coordination is needed among the agents. For example, intelligent lighting, which maintains illumination constraints while minimizing power usage, is one application where the distributed algorithm can be applied directly.
  • Defense presentation (powerdot slides): "Engineering Serendipity: Design and Analysis of Optimal Task-Processing Agents" (PDF [LARGE 6 MB]) (source code repository)
      Abstract:
      Biomimicry has been a great source of novel ideas for bad technological design. Both engineers and biologists look for concrete examples where behavioral models can be applied directly. Consequently, engineers overlook important theoretical assumptions behind those models, and biologists overlook non-traditional design problems because of the attraction to the charismatic megafauna of technology (e.g., robots). In this talk, efforts to create general optimal agent-based frameworks that can be specialized for different disciplines are described.

      First, a solitary task-processing agent framework is presented that unifies behavioral models from engineering design and biological analysis. Using this framework, an example decision-making algorithm for autonomous vehicles is generated that performs better than the conventional bio-inspired algorithm. Because both algorithms fit within the unified framework, they are computationally similar, and autonomous vehicles using the bio-inspired algorithm can be easily modified to use the better performing algorithm. Furthermore, ostensibly irrational observed animal behaviors that vary from the predictions of classical optimal foraging theory are shown to be optimal within the generalized task-processing framework.

      Finally, the talk presents two frameworks for collective task processing in groups of agents. First, a generic model of asynchronous distributed cooperative task processing meant to mimic human, non-human, and artificial systems is presented. Despite local utility maximization at each agent, patterns emerge from the group that are qualitatively similar to load balancing. Second, a distributed solver for constrained optimization is presented that has applications in power generation, intelligent-light design, and eusocial-insect-foraging analysis.


National Science Foundation (NSF) Graduate Teaching Fellows in K-12 Education (GK-12)

End of Year Overview
  • "Selling Your Soul for Science: Notes on Being an NSF GK-12 Fellow" (PPT)



 


appalling appalling
appalling appalling
email me email me
 
569210 hits
(200 today)
 
Terms of Use
Ted Pavlic <ted@tedpavlic.com>   appalling appalling appalling appalling email me email me GPG Public Key: D/L, View, Ubuntu, MIT, PGP (verified) (ID: E1E66F7C) This Page Last Updated on Monday, June 13, 2011, 5:00 pm GMT