ArticlePDF Available

The Real-World Navigator

Authors:
Marko Balabanovic at Digital Catapult, London, United Kingdom
Marko Balabanovic
  • Digital Catapult, London, United Kingdom
Craig Becker
Craig Becker
Craig Becker
  • This person is not on ResearchGate, or hasn't claimed this research yet.
Sarah K. Morse
Sarah K. Morse
Sarah K. Morse
  • This person is not on ResearchGate, or hasn't claimed this research yet.
Illah R. Nourbakhsh
Illah R. Nourbakhsh
Illah R. Nourbakhsh
  • This person is not on ResearchGate, or hasn't claimed this research yet.
Download full-text PDF
Read full-text
Download citation

Abstract

The success of every mobile robot application hinges on the ability to navigate robustly in the real world. The problem of robust navigation is separable from the challenges faced by any particular robot application. We offer the Real-World Navigator as a solution architecture that includes a path planner, a map-based localizer, and a motion control loop that combines reactive avoidance modules with deliberate goal-based motion. Our architecture achieves a high degree of reliability by maintaining and reasoning about an explicit description of positional uncertainty. We provide two implementations of real-world robot systems that incorporate the Real-World Navigator. The Vagabond Project culminated in a robot that successfully navigated a portion of the Stanford University campus. The Scimmer project developed successful entries for the AAAI 1993 Robotics Competition, placing first in one of the two contests entered. 1 Introduction Current research on autonomous mobile robots has high...
Content uploaded by Marko Balabanovic
Author content
All content in this area was uploaded by Marko Balabanovic on Jul 30, 2015
Content may be subject to copyright.
Map
Task Level
Navigator Level
Robot Level
(x,y)
goals
‘success’
’impossible’
’lost’
motion
primitives
sensor
data
Path
Planner
Control
Loop
Localizer
Map
motion
primitives
sensor
data
’success’
’impossible’
’lost’
(x,y)
goals
Navigator Level
... If the sensors are assumed to be imperfect, how is the uncertainty dealt with? In one example [1], a region of uncertainty is maintained such that the robot is guaranteed to be within that region at all times. The region of uncertainty is expanded as appropriate and contracted when landmarks are available which the robot can use to localize itself. ...
... The Real-World Navigator [1] architecture consists of three levels. The Task level keeps track of the robot's goal locations. ...
... This allows the slow planner to run without affecting the response of the reactive component of the architecture. Explicit reasoning about deadlines [7] and explicit maintenance of a region of uncertainty [1] can be more easily implemented in Gat's framework since the addition of these high-level tasks does not slow down the control loop; however, it remains to be seen how much can be added to the planner's tasks before it becomes so slow that it is unable to ever provide meaningful, up-to-date advice. ...
A Survey of Robot Architectures
Article
  • Aug 1999
The goal of a robot system in general is to solve problems of navigation and/or manipulation. To construct a robot system, one must answer many questions regarding the environment in which the robot is expected to operate and the robot architecture best suited for the problem. In this paper, we look at six different approaches to building robot systems. We examine the assumptions they make regarding the environment and themselves, determine whether they are justified, and suggest extensions to these robots. We compare these approaches and draw some conclusions about robot architectures in general. 1. Introduction The goal of a robot system in general is to solve problems of navigation and/or manipulation. To construct a robot system, one must determine the nature of the environment in which the robot is expected to operate and the exact problem(s) that need to be solved. On that basis, the robot architecture can be designed and the robot constructed. We will examine six approaches to ...
View
SSS: A hybrid architecture applied to robot navigation
Conference Paper
Full-text available
  • Jun 1992
Describes a three-layer architecture, SSS, for robot control. It combines a servo-control layer, a subsumption layer, and a symbolic layer in a way that allows the advantages of each technique to be fully exploited. The key to this synergy is the interface between the individual subsystems. The design of situation recognizers that bridge the gap between the servo and subsumption layers, and event detectors that link the subsumption layers and symbolic layers are discussed. The development of such a combined system is illustrated by a fully implemented indoor navigation example. The resulting robot was able to automatically map office building environments, and smoothly navigate through them at the rapid speed of 2.6 feet per second
View
Plans and Behavior in Intelligent Agents
Article
Full-text available
  • Feb 1970
An "intelligent agent" is a versatile and adaptive computer system. It performs diverse cognitive and physical behaviors in its efforts to achieve multiple goals in a dynamic, uncertain task environment. In this paper, we address the question: How should plans influence the cognitive and physical behavior of intelligent agents? In contrast to the well-known model of plans as executable programs, we propose that intelligent agents can make better use of plans that describe their intended behavior. This kind of model has been discussed by other researchers, but it has not been operationalized. We instantiate this model with operational definitions of plans, planning, and plan following. We present an agent architecture for using such plans to guide an agent's cognitive and physical behaviors. Experimental results from an office robot demonstrate important capabilities engendered by our approach: (a) coordination of diverse cognitive and physical behaviors to achieve multiple goals; (b) r...
View
Robot Motion Planning
Book
  • Jan 1991
1 Introduction and Overview.- 2 Configuration Space of a Rigid Object.- 3 Obstacles in Configuration Space.- 4 Roadmap Methods.- 5 Exact Cell Decomposition.- 6 Approximate Cell Decomposition.- 7 Potential Field Methods.- 8 Multiple Moving Objects.- 9 Kinematic Constraints.- 10 Dealing with Uncertainty.- 11 Movable Objects.- Prospects.- Appendix A Basic Mathematics.- Appendix B Computational Complexity.- Appendix C Graph Searching.- Appendix D Sweep-Line Algorithm.- References.
View
Landmark-based robot motion planning
Article
  • Jan 1991
  • Lect Notes Comput Sci
This paper describes a reduced version of the gen-eral motion planning problem in the presence of uncertainty and a complete polynomial algorithm solving it. This algorithm computes a guaran-teed plan by backchaining non-directional preim-ages of the goal until one fully contains the set of possible initial positions of the robot. It assumes that landmarks are scattered across the workspace, that robot control and sensing are perfect within the fields of influence of these landmarks, and that control is imperfect and sensing null outside these fields. The satisfaction of these assumptions may require the robot and/or its workspace to be specifically engineered. This leads us to view robot/workspace engineering as a means to make planning problems tractable. The algorithm was implemented in an operational planner and run on many examples. Non-implemented extensions of the planner are also discussed.
View