Abstractions and security concepts for a robot operating system
Finnicum, Murph
Content Files
Murph_Finnicum.pdf
Loading…
Download Files
Loading…
Download Counts (All Files)
Loading…
Edit File
Loading…
Permalink
https://hdl.handle.net/2142/45567
Description
Title
Abstractions and security concepts for a robot operating system
Author(s)
Finnicum, Murph
Issue Date
2013-08-22T16:48:03Z
Director of Research (if dissertation) or Advisor (if thesis)
King, Samuel T.
Department of Study
Computer Science
Discipline
Computer Science
Degree Granting Institution
University of Illinois at Urbana-Champaign
Degree Name
M.S.
Degree Level
Thesis
Date of Ingest
2013-08-22T16:48:03Z
Keyword(s)
Robot Operating System
Robot Applications
Abstract
As general purpose robots begin to find their way into the household and
workplace, there will be a demand for software to run on these robots. My
research group forsees the proliferation of robot apps that use a common set
of abstractions to allow them to function on a variety of hardware platforms.
In this thesis, I introduce a robot operating system to support these apps
and detail the abstractions that it provides. I present many lessons learned
from developing and debugging a number of such apps, and discuss a novel
concept wherein apps and libraries are allowed to seek help from outside
sources when they are unable to accomplish their goals. I show that our
framework allows a robot to effectivly deal with challenges, such as user
authentication. I demonstrate a simple bartender app to fetch drink orders
for students, and it is succesfully able to deliver them in 10/10 trials in real-
world conditions. I also present CLASS, a new system capable of identifying
users robustly. I propose a framework for integrating a wide range of sensor
values into an algorithm for identifying users, even if an attacker actively tries
to impersonate a user. Our system is evaluated by using our reference robot
and this platform to build a robot application that buys coffee at our local
coffee shop for a user, without requiring explicit authentication. I evaluate
CLASS under an adversarial model experimentally and find it to be robust
and resilient to attack.
Use this login method if you
don't
have an
@illinois.edu
email address.
(Oops, I do have one)
IDEALS migrated to a new platform on June 23, 2022. If you created
your account prior to this date, you will have to reset your password
using the forgot-password link below.
Murph_Finnicum.pdf
