E. Tunstel's Publication Abstracts
54
53
52
51
50
49
48
47
46
45
44
43
42
41
40
39
38
37
36
35
34
33
32
31
30
29
28
27
26
25
24
23
22
21
20
19
18
17
16
15
14
13
12
11
10
9
8
7
6
5
4
3
2
1
Back to publications list
Home
54: Autonomous Mobility Software Validation
Challenges for Planetary Surface Missions
This paper discusses characteristics of the surface mobility problem
that present challenges for validation of autonomous mobility software.
We provide a general description of a commonly used hardware-centric and
dynamic approach to mobility software validation for flight systems.
We also highlight some effects and impacts of flight schedules and
deliveries on this validation approach and related mobility software
development. Finally, we offer suggestions for dealing with some of
the challenges and improving the process for autonomous mobility software
development and validation on future projects.
53: Planetary Rover Developments Supporting
Mars Exploration, Sample Return and Future Human-Robotic Colonization
We overview our recent research on planetary mobility. Products of this effort
include the Field Integrated Design & Operations rover (FIDO), Sample Return
Rover (SRR), reconfigurable rover units that function as an All Terrain
Explorer (ATE), and a multi-Robot Work Crew of closely cooperating rovers
(RWC). FIDO rover is an advanced technology prototype; its design and field
testing support NASA's development of long range, in situ Mars surface science
missions. Complementing this, SRR implements autonomous visual recognition,
navigation, rendezvous, and manipulation functions enabling small object
pick-up, handling, and precision terminal docking to a Mars ascent vehicle for
future Mars Sample Return. ATE implements on-board reconfiguration of rover
geometry and control for adaptive response to adverse and changing terrain,
e.g., traversal of steep, sandy slopes. RWC implements coordinated control of
two rovers under closed loop kinematics and force constraints, e.g.,
transport of large payloads, as would occur in robotic colonies at future
Mars outposts. RWC is based in a new extensible architecture for
decentralized control of, and collective state estimation by multiple
heterogeneous robotic platforms -- CAMPOUT; we overview the key architectural
features. We have conducted experiments with all these new rover system
concepts over variable natural terrain. For each of the above developments,
we summarize our approach, some of our key experimental results to date,
and our future directions of planned development.
52: Applied Soft Computing Strategies for Autonomous Field Robotics
This chapter addresses computing strategies designed to enable field mobile
robots to execute tasks requiring effective autonomous traversal of natural
outdoor terrain. The primary focus is on computer vision-based perception and
autonomous control. Hard computing methods are combined with applied soft
computing strategies in the context of three case studies associated with
real-world robotics tasks including planetary surface exploration and
land survey/reconnaissance. Each case study covers strategies implemented
on wheeled robot research prototypes designed for field operations.
51: Approximate Reasoning for Safety and Survivability of Planetary Rovers
Operational safety and health monitoring are critical matters for
autonomous field mobile robots such as planetary rovers operating
on remote and challenging terrain. This paper describes relevant
rover safety and health issues and presents a soft computing approach
to maintaining vehicle safety in a navigational context. The proposed
rover safety module is composed of two distinct behaviors: safe attitude
(pitch and roll) management and safe traction management. Fuzzy logic
approaches to approximate reasoning about safe attitude and traction
management on outdoor terrain are presented. Sensing of vehicle safety
status coupled with visual neural network-based perception of terrain
quality are used to infer safe speeds during rover traversal. In
addition, an approximate reasoning approach to self-regulation of
internal operating condition is briefly discussed. The core theoretical
foundations of the applied soft computing techniques is presented and
supported by descriptions of field tests and laboratory experimental
results. For autonomous rovers, the approach provides an intrinsic
safety cognizance and a capacity for reactive mitigation of
navigation risks.
50: Rule-Based Reasoning and Neural Network
Perception for Safe Off-Road Mobility
Operational safety and health monitoring are critical matters for
autonomous field mobile robots such as planetary rovers operating on
challenging terrain. This paper describes relevant rover safety and
health issues and presents an approach to maintaining vehicle safety
in a mobility and navigation context. The proposed rover safety module
is composed of two distinct components: safe attitude (pitch and roll)
management and safe traction management. Fuzzy logic approaches to
reasoning about safe attitude and traction management are presented,
wherein inertial sensing of safety status and vision-based neural
network perception of terrain quality are used to infer safe speeds
of traversal. Results of initial field tests and laboratory experiments
are also described. The approach provides an intrinsic safety cognizance
and a capacity for reactive mitigation of robot mobility and navigation risks.
49: Sensing and Perception Challenges in Planetary Surface Robotics
This expository paper describes sensing and perception issues facing
the space robotics community concerned with deploying autonomous rovers
on other planetary surfaces. Challenging sensing problems associated
with rover surface navigation and manipulation functions are discussed
for which practical solutions from sensor developers would vastly
improve rover capabilities. Some practical concerns that impact sensor
selection based on mass, volume, power, and operability constraints
are also discussed. The intent is to present challenges to facilitate
alignment of new sensing solutions with key sensing requirements of
planetary surface robotics.
48: FIDO Rover Field Trials as Rehearsal for the 2003 Mars Exploration Rover Mission
This paper describes recent extended field trials performed by NASA-JPL
using the FIDO (Field Integrated Design & Operations) rover, an advanced
technology development platform and research prototype for the next Mars
rover mission planned by NASA. A realistic physical simulation of the
NASA 2003 Mars Exploration Rovers mission was achieved through
collaborative efforts of roboticists, planetary scientists, and
mission operations personnel. An overview of the objectives, approach,
and results is reported.
47: FIDO Rover System
Enhancements for High-Fidelity Mission Simulations
This paper describes recent field experimentation and testing performed
by NASA-JPL using the FIDO (Field Integrated Design & Operations) rover,
an advanced technology development platform and research prototype for
the next Mars rover mission planned by NASA. System enhancements are
described in particular, that served to improve the utility of the FIDO
rover as a platform for high-fidelity, physical simulation of a Mars
surface mission in terrestrial analogue environments. Field testing
activities aimed at simulation of the 2003 NASA Mars Exploration Rover
mission in cooperation with mission science and engineering personnel
are reported.
46: Rover Autonomy for Long Range
Navigation and Science Data Acquisition on Planetary Surfaces
This paper describes recent work undertaken at the Jet Propulsion
Laboratory in Pasadena, CA in the area of increased rover
autonomy for planetary surface operations. The primary vehicle for
this work is the Field Integrated, Design and Operations (FIDO)
rover. The FIDO rover is an advanced technology prototype that is
a terrestrial analog of the Mars Exploration Rovers (MER) being
sent to Mars in 2003. We address the autonomy issue through
improved integration of rover based sensing and higher level
onboard planning capabilities. The sensors include an inertial
navigation unit (INU) with 3D gyros and accelerometers, a sun
sensor, mast and body mounted imagery, and wheel encoders.
Multisensor fusion using an Extended Kalman Filter (EKF)
approach coupled with pattern recognition and tracking algorithms
has enabled the autonomy that is necessary for maximizing science
data return while minimizing the number of ground loop
interactions. These algorithms are coupled with a long range
navigation algorithm called ROAMAN (Road Map Navigation) for
an integrated approach to rover autonomy. We also report the
results of algorithm validation studies in remote field trials at
Black Rock Summit in Central Nevada, California's Mojave Desert,
and the Arroyo Seco at JPL.
45: Fuzzy Behavior Hierarchies for Multi-Robot Control
Hierarchical approaches and methodologies are commonly used for control
system design and synthesis. Well-known model-based techniques are
often applied to solve problems of complex and large-scale control systems.
The general philosophy of decomposing control problems into modular and more
manageable subsystem control problems applies equally to the growing domain
of intelligent and autonomous systems. However, for this class of systems new
techniques for subsystem coordination and overall system control are often
required. This paper presents an approach to hierarchical control design and
synthesis for the case where the collection of subsystems is comprised of
fuzzy logic controllers and fuzzy knowledge-based decision systems. The
approach is used to implement hierarchical behavior-based controllers for
autonomous navigation of one or more mobile robots. Theoretical details of
the approach are presented, followed by discussions of practical design and
implementation issues. Example implementations realized on various physical
mobile robots are described to demonstrate how the techniques may be applied
in practical applications involving homogeneous and heterogeneous robot teams.
44: A Distributed Diode Laser Spectrometer
for Mapping Biogenic Gases on the Martian Surface
A sensitive versatile diode laser spectrometer is being designed,
constructed and tested that can map biogenic gas levels on the Martian surface.
Gases to be measured include water vapor, methane, oxygen, ammonia and hydrogen
sulfide. The unique feature of the spectrometer is that the diode laser
source and the laser beam detector are mounted on separate, semi-autonomous
rovers. With this distributed spectrometer, theportion of atmosphere sampled
is line of sight between the two rovers, allowing for surveying over large
distances or over small distances, efficiently mapping the target terrain or
geological feature for these gases.
Diode lasers are ideal sensors for space missions. They occupy less than a
cubic centimeter and weigh about two grams. The photodiode detectors are
similar in size and mass and do not require a power supply.
The rover mapping system presents several challenges atypical of normal rover
navigation problems, including regulation and adjustment of instrument
path length, dual rover hazard avoidance and subsequent recovery of search
path tracking, as well as handling intermittent data acquisition due to
interruption of the laser beam by terrain obstructions. The approach taken
for simplifying the mapping process involves development of a local
coordinate system. A two-robot based system of landmark recognition is being
explored for robot localization. The vision system employed in the
alignment system will be used to determine the relative position of the
companion robot in three dimensions.
The vision system with its pan and tilt mechanism has been built and
tested to twenty meters, the diode laser system has been built and laboratory
spectra obtained for ambient water vapor over an open path of one meter.
The final construction of a prototype system to be used for field tests is
being completed. Plans for testing in the Canadian Arctic are being pursued.
43: Ethology as an Inspiration for Adaptive Behavior Synthesis in Autonomous Planetary Rovers
A navigation and control approach that supports
adaptive behavior in rovers is presented. It is motivated by ethological
models that suggest hierarchical organizations of behavior. The methodology
employs fuzzy logic as a means to emulate animal behavior control
mechanisms such as behavior activation levels, multi-behavior
modulation, and threshold activation.
The paper describes how these concepts can be tailored for
autonomous navigation to provide a suitable framework for situated
adaptation in rover control algorithms. In addition, an interesting
characteristic of observed behavioral interactions achieved by an
implementation of the approach is discussed, which is analogous to
phenomena observed in measurements of animal brain activity during
transitions between distinct behaviors.
42: Robotic Automation for Space:
Planetary surface exploration, terrain-adaptive mobility,
and multi-robot cooperative tasks
During the last decade, there has been significant progress toward a
supervised autonomous robotic capability for remotely controlled scientific
exploration of planetary surfaces. While planetary exploration potentially
encompasses many elements ranging from orbital remote sensing to subsurface
drilling, the surface robotics element is particularly important to advancing
in situ science objectives. Surface activities include a direct
characterization of geology, mineralogy, atmosphere and other
descriptors of current and historical planetary processes -- and ultimately
-- the return of pristine samples to Earth for detailed analysis.
Toward these ends, we have conducted a broad program of research on
robotic systems for scientific exploration of the Mars surface, with minimal
remote intervention. The goal is to enable high productivity semi-autonomous
science operations where available mission time is concentrated on robotic
operations, rather than up-and-down-link delays. Results of our work include
prototypes for landed manipulators, long-ranging science rovers,
sampling/sample return mobility systems, and more recently, terrain-adaptive
reconfigurable/modular robots and closely cooperating multiple rover systems.
The last of these are intended to facilitate deployment of planetary
robotic outposts for an eventual human-robot sustained scientific presence.
We overview our progress in these related areas of planetary robotics R&D,
spanning 1995-to-present.
41:Enhancing Fuzzy Robot Navigation Systems by Mimicking
Human Visual Perception of Natural Terrain Traversability
This paper presents a technique for learning to assess terrain traversability
for outdoor mobile robot navigation using human-embedded logic and real-time
perception of terrain features extracted from image data. The methodology
utilizes a fuzzy logic framework and vision algorithms for analysis of the
terrain. The terrain assessment and learning methodology is tested and
validated with a set of real-world image data acquired by an onboard
vision system.
40: Terrain-based Navigation of Mobile Robots:
A fuzzy logic approach
This paper presents a new strategy for autonomous
navigation of field mobile robots on hazardous natural
terrain using a fuzzy logic approach and a novel measure
of terrain traversability. The navigation strategy
is comprised of three simple, independent behaviors:
seek-goal, traverse-terrain, and avoid-obstacle. The
recommendations from these three behaviors are combined
through appropriate weighting factors to generate the
final steering and speed commands that are executed by the robot.
The weighting factors are produced by fuzzy logic
rules that take into account the current status of the robot.
This navigation strategy requires no a priori information
about the environment, and uses the on-board traversability
analysis to enable the robot to select relatively easy-to-traverse
paths autonomously. Field test results obtained from implementation
of the proposed algorithms on the commercial Pioneer AT rover are
presented. These results demonstrate the real-time capabilities of
the terrain assessment and fuzzy logic navigation algorithms.
39:A Rule-Based Fuzzy Traversability Index for
Mobile Robot Navigation
This paper presents a fuzzy logic rule-based approach to assessing terrain
quality for autonomous robot navigation. Using real-time measurements of terrain
characteristics retrieved from imagery data, the suitability of the terrain
for traversal is represented by a rule-based Fuzzy Traversability Index. These
characteristics include, but are not limited to: terrain slope, roughness,
hardness, and discontinuity. The representation incorporates an intuitive,
linguistic approach for expressing terrain characteristics that is robust
with respect to imprecision and uncertainties in the terrain measurements.
The terrain assessment methodology is tested and validated with a set of
real-world imagery data. These experiments demonstrate the capability of the
terrain classification algorithm for perceiving mobility hazards associated
with terrain traversal.
38:Fuzzy Rule-Based Reasoning for Rover Safety
and Survivability
Operational safety and health monitoring are critical matters for autonomous
field mobile robots such as planetary rovers operating on challenging terrain.
This paper describes relevant rover safety and health issues and presents an
approach to maintaining vehicle safety in a navigational context. The
proposed rover safety module is composed of two distinct components: safe
attitude (pitch and roll) management and safe traction management. Fuzzy
logic approaches to reasoning about safe attitude and traction management
are presented, wherein sensing of safety status and perception of terrain
quality are used to infer safe speeds of traversal. Results of field tests
and laboratory experiments are also described. The approach provides an
intrinsic safety cognizance and a capacity for reactive mitigation of
navigation risks
37:Genetic and Evolutionary Methods for Mobile
Robot Motion Control and Path Planning
A variety of evolutionary algorithms, operating according to Darwinian concepts,
have been proposed to solve problems of common engineering applications.
Applications often involve automatic learning of nonlinear mappings that
govern the behavior of control systems, as well as parallel search strategies
for solving multiobjective optimization problems. In many cases, hybrid
applications of soft computing methods have proven to be effective in designing
intelligent control systems. This chapter presents two instances of such hybrid
applications to problems of mobile robot control. In particular, evolutionary
computation and fuzzy logic are combined to solve robot motion control and
path planning problems. The first part of the chapter describes a methodology
for applying genetic programming (GP) to design a fuzzy logic steering controller
for mobile robot path tracking. Genetic programming is employed to learn the
rules and membership functions of the fuzzy logic controller, and also to handle
selection of fuzzy set intersection operators (t-norms). The second part of
the chapter describes an application of fuzzy logic to enhance the performance
of an evolutionary robot path planning system. In this case, fuzzy logic is
employed in the selection phase of the simulated evolution process.
36:Soft Computing Approach to Safe Navigation of
Autonomous Planetary Rovers
As an integral part of its initiatives to explore the planet Mars, NASA has
opted to employ mobile robots that are designed to rove across the surface
in search of clues and evidence about the geologic and climatic history of
the planet. In the summer of 1997 NASA deployed an autonomous rover, named
Sojourner, on Mars. In 2003, NASA plans to launch a Mars mission that will
use two rovers to explore distinct regions of the planet1s surface, each
having greater mobility and autonomy than Sojourner. In order to advance
rover navigation capabilities beyond those of Sojourner, and even the twin
Mars exploration rovers planned for the 2003 mission, advanced algorithms and
computational approaches to autonomy and intelligent control must be pursued
that comply with practical constraints. This chapter describes fundamental
research aimed at achieving objectives of long-range mobility and increased
autonomy through application of soft computing techniques for safe and
reliable autonomous rover navigation. A fuzzy-logic-based reasoning and
control framework is presented that is complemented by neural networks and
visual perception algorithms to realize a practical rover navigation system.
A discussion of experimental investigations using a commercial mobile robot
as a testbed for outdoor navigation research is provided.
35:Fuzzy-Behavior Synthesis, Coordination, and Evolution
in an Adaptive Behavior Hierarchy
Autonomous controllers with adaptive behavioral capabilities must
accommodate uncertainty. Fuzzy logic is particularly useful for this
purpose. A behavior-based fuzzy control approach to autonomous local
navigation is presented here. A hierarchical architecture is described
which consists of multiple sets of fuzzy rules. Each rule set represents
a decision or motion behavior that contributes to a dynamic map from
stimuli to goal-oriented actions. Synthesis of a behavior hierarchy which
combines primitive motion behaviors to produce navigation behaviors is
discussed. This is coupled with a technique for multi-behavior coordination
for which fuzzy rule learning by genetic programming is applied. The overall
approach is verified by experimental results in an indoor environment without
use of explicit maps. Performance is demonstrated on a mobile robot with
significant mechanical imperfections. Descriptions of the experiments,
apparatus, and setting are included.
34:Remote Surface Exploration with Multiple
Soft Computing Based Cooperative Rovers
Current research at the ACE Center in the use of fuzzy behavior-based control,
coupled with a neural network approach to sensor fusion, for cooperative
robotics is described using a remote planetary surface exploration scenario.
The proposed architecture is an extension, to multiple mobile robots, of the
hierarchical fuzzy logic controller developed earlier by one of the authors.
Robots are dynamically organized, based on inter-robot communication range,
into a coordinator-scouts hierarchy. Preliminary results from simulation
runs are presented and appreciated.
33:Distributed Neuro-Evolutionary Path Planning
Recently there has been a growing interest in the use path planners
based on evolutionary computation. To date, a number of single agent
evolutionary path planners have been proposed. In real-world situations,
it is possible for single agent path planners to become overwhelmed by
the sheer volume of information needed to be processed in order to
develop paths in real-time. In this paper, two distributed evolutionary
computations (DECs) are presented and compared on ten randomly generated
path planning problems. In theses DECs, the variables that make up the
structure of candidate solutions are dustributed among the path planning
agents rather than the population. By co-evolving radial basis function
neural networks, these DECs are able to dramatically reduce the number
of nodes needed to represent candidate paths and lead to smooth realistic
paths from source to destination.
32: Genetic Programming Design
of Fuzzy Controllers for Mobile Robot Path Tracking
Genetic programming (GP) is an evolutionary strategy that attempts to deal
with the notion of how computers can learn to solve problems without being
explicitly programmed. It has been demonstrated that GP, under the
influence of Darwinian concepts, could genetically breed computer
programs to approximately solve problems in a variety of applications.
One primary example is its application to the problem of automatically
learning nonlinear mappings that govern the behavior of control systems.
It is demonstrated here that GP can formulate such nonlinear maps in the
form of fuzzy control rules, which yield comparable or better performance
than one derived through manual design using trial-and-error. The
objective is to address the efficient implementation of GP for the
discovery of knowledge bases intended for use in fuzzy logic controller
applications. Efficiency is achieved with a C programming language
implementation of GP, which is applied to a mobile robot steering control
problem. Robot path following performance is compared to results obtained
using an existing GP implementation in the LISP programming language.
It is demonstrated that the C implementation has a definite advantage
with regard to computational speed of evolution. In this work, we have
extended the application of GP to handle simultaneous evolution of
membership functions and rule bases for the same control problem.
Furthermore, GP is used to handle selection of fuzzy t-norms. It is
concluded that simultaneous evolution of rule bases and membership
functions with t-norm selection results in enhanced performance of
the evolved controllers. Finally, the robustness characteristics of
the genetically evolved fuzzy controllers are investigated by examining
the effects of sensor measurement noise and an increase in the robot's
nominal forward velocity.
31:Soft Computing for Autonomous Robotic
Systems
Neural networks (NN), genetic algorithms (GA), and genetic programming (GP)
are augmented with fuzzy logic-based schemes to enhance artificial
intelligence of automated systems. Such hybrid combinations exhibit added
reasoning, adaptation, and learning ability. In this expository article, three
dominant hybrid approaches to intelligent control are experimentally applied
to address various robotic control issues which are currently under
investigation at the NASA Center for Autonomous Control Engineering.
The hybrid controllers consist of a hierarchical NN-fuzzy controller applied
to a direct drive motor, a GA-fuzzy hierarchical controller applied to
position control of a flexible robot link, and a GP-fuzzy behavior based
controller applied to a mobile robot navigation task. Various strong
characteristics of each of these hybrid combinations are discussed and
utilized in these control architectures. The NN-fuzzy architecture takes
advantage of NN for handling complex data patterns, the GA-fuzzy architecture
utilizes the ability of GA to optimize parameters of membership functions for
improved system response, and the GP-fuzzy architecture utilizes the symbolic
manipulation capability of GP to evolve fuzzy rule-sets.
30:Soft Computing-based Design and Control
for Mobile Robot Path Tracking
A variety of evolutionary algorithms, operating according to Darwinian
concepts, have been proposed to approximately solve problems of common
engineering applications. Increasingly common applications involve
automatic learning of nonlinear mappings that govern the behavior of
control systems. In many cases where robot control is of primary
concern, the systems used to demonstrate the effectiveness of
evolutionary algorithms often do not represent practical robotic
systems. In this paper, genetic programming (GP) is the evolutionary
strategy of interest. It is applied to learn fuzzy control rules for
a practical autonomous vehicle steering control problem, namely, path
tracking. GP handles the simultaneous evolution of membership functions
and rule bases for the fuzzy path tracker. As a matter of practicality,
robustness of the genetically evolved fuzzy controller is demonstrated
by examining the effects of sensor measurement noise and an increase in
the robot's nominal forward velocity.
29:Genetic Programming of Full Knowledge Bases for Fuzzy Logic Controllers
Genetic programming (GP) is applied to automatic discovery of full
knowledge bases for use in fuzzy logic control applications. An
extension to a rule learning GP system is presented that achieves
this objective. In addition, GP is employed to handle selection of
fuzzy set intersection operators (t-norms). The new GP system is
applied to design a mobile robot path tracking controller and
performance is shown to be comparable to that of a manually designed
controller.
28:Fuzzy Behavior Modulation with Threshold Activation for Autonomous Vehicle Navigation
This paper decribes fuzzy logic techniques used in a hierarchical
behavior-based architecture for robot navigation. An architectural
feature for threshold activation of fuzzy-behaviors is emphasized,
which is potentially useful for tuning navigation performance in
real world applications. The target application is autonomous
local navigation of a small planetary rover. Threshold activation
of low-level navigation behaviors is the primary focus. A preliminary
assessment of its impact on local navigation performance is provided
based on computer simulations.
27:An Integrated Architecture for Cooperating Rovers
This paper presents a rover execution architecture for controlling multiple,
cooperating rovers. The overall goal of this architecture is to coordinate
multiple rovers in performing complex tasks for planetary science. This
architecture integrates a number of systems and research efforts on single
rovers and extends them for multiple rover operations. Techniques from a
number of different fields are utilized, including AI planning and scheduling,
real-time systems and simulation, terrain modeling, and AI machine learning.
In this paper, we discuss each architecture component, describe how components
interact and present the geological scenario we are using to evaluate the
overall architecture.
26:Evolution of Autonomous Self-Righting Behaviors
for Articulated Nanorovers
Miniature rovers with articulated mobility mechanisms are being developed
for planetary surface exploration on Mars and small solar system bodies.
These vehicles are designed to be capable of autonomous recovery from
overturning during surface operations. This paper describes a proposed
computational means of developing motion behaviors that achieve the
autonomous recovery function. Its aim is to reduce the effort involved
in developing self-righting control behaviors. The approach is based
on the integration of evolutionary computing with a dynamics simulation
environment for evolving and evaluating motion behaviors. The automated
behavior design approach is outlined and its underlying genetic programming
infrastructure is described.
25: Embedded Control of a Miniature Science Rover
for Planetary Exploration
Recent advances in micro-technology and mobile robotics have enabled the
development of extremely small, automated vehicles for new application
frontiers. One of these possible applications is the use of miniature robotic
vehicles equipped with on-board science instruments for planetary surface
exploration. One such vehicle is being developed as part of a technology
research task at the Jet Propulsion Laboratory. A rover prototype has been
integrated into a package of several hundred grams in mass. Aspects of the
embedded rover control and software implementation are discussed which
enable mobility and operation of science instruments for navigation and
surface exploration.
24: Soft Computing Paradigms for Hybrid Fuzzy
Controllers: Experiments and applications
Neural Networks (NN), Genetic Algorithms (GA), and Genetic Programs (GP)
are often augmented with fuzzy logic-based schemes to enhance artificial
intelligence of a given system. Such hybrid combinations are expected to
exhibit added intelligence, adaptation, and learning ability. In this paper,
implementation of three hybrid fuzzy controllers are discussed and
verified by experimental results. These hybrid controllers consist of a
hierarchical NN-fuzzy controller applied to a direct drive motor, a
GA-fuzzy hierarchical controller applied to a flexible robot link, and
a GP-fuzzy behavior-based controller applied to a mobile robot
navigation task. It is experimentally shown that all three architectures are
capable of significantly improving the system response.
23: Multiobjective Evolutionary Path Planning via
Fuzzy Tournament Selection
This paper introduces a new selection algorithm that can be used for
evolutionary path planning systems. This new selection algorithm combines
fuzzy inference along with tournament selection to select candidate paths
(CPs) to be parents based on: (1) the Euclidean distance from origin to
destination, (2) the sum of the changes in the slope of a path, and (3) the
average change in the slope of a path. In this paper, we provide a
detailed description of the fuzzy inference system used in the new fuzzy
tournament selection algorithm (FTSA) as well as some examples of its
usefulness. We use 12 instances of our FTSA to rank a population of CPs
using the above criteria. Based on its path ranking capability, we show
how the FTSA can obviate the need for the development of an explicit
multiobjective evaluation function. Finally, we use the FTSA to enhance
the performance of an existing evolutionary path planning system called
GEPOA.
22: Autonomous Navigation using an Adaptive Hierarchy
of Multiple Fuzzy-Behaviors
Adaptive behavioral capabilities are necessary for robust mobile robot
navigation in non-engineered environments. Robust behavior requires
that uncertainty be accommodated in the robot control system, especially
when autonomy is desired. Fuzzy logic control technology enables development
of controllers which can provide the necessary computational intelligence
in real-time. This paper presents the incorporation of fuzzy logic, into the
framework of behavior-based control. An architecture for hierarchical
behavior control is presented in which control decisions result from a
consensus of recommendations offered only by behaviors that are applicable
to current situations. Aspects of multiple fuzzy-behavior coordination are
discussed with application autonomous navigation without explicit maps.
Performance and robustness is demonstrated by implementation on a mobile
robot with significant mechanical imperfections.
21: Behavior Hierarchy for Autonomous Mobile Robots:
Fuzzy-behavior modulation and evolution
Realization of autonomous behavior in mobile robots, using fuzzy logic control,
requires formulation of rules which are collectively responsible for necessary
levels of intelligence. Such a collection of rules can be conveniently
decomposed and efficiently implemented as a hierarchy of fuzzy-behaviors.
This article describes how this can be done using a behavior-based
architecture. A behavior hierarchy and mechanisms of control decision-making
are described. In addition, an approach to behavior coordination is
described with emphasis on evolution of fuzzy coordination rules using the
genetic programming (GP) paradigm. Both conventional GP and steady-state GP
are applied to evolve a fuzzy-behavior for sensor-based goal-seeking.
The usefulness of the behavior hierarchy, and partial design by GP,
is evident in performance results of simulated autonomous navigation.
20: Adaptive Fuzzy-Behavior Hierarchy for Autonomous Navigation
Adaptive behavioral capabilities are necessary for robust navigation in
non-engineered environments. A control approach to adaptive behavior is
described which exploits the approximate reasoning facility of fuzzy logic.
In particular, a behavior-based architecture for hierarchical fuzzy control
of mobile robots is presented. Its structure is described as well as
mechanisms of control decision-making which give rise to adaptive behavior.
Control decisions result from a consensus of recommendations offered only by
behaviors that are applicable to current situations. Indoor navigation
examples demonstrate the practicality of the approach and reveals
characteristics of multiple behavior interaction.
19: Fuzzy Behavior-based Navigation for Planetary
Microrovers
Adaptive behavioral capabilities are necessary for robust rover navigation
in unstructured and partially-mapped environments. A control approach is
described which exploits the approximate reasoning capability of fuzzy
logic to produce adaptive motion behavior. In particular, a behavior-based
architecture for hierarchical fuzzy control of microrovers is presented. Its
structure is described, as well as mechanisms of control decision-making
which give rise to adaptive behavior. Control decisions for local navigation
result from a consensus of recommendations offered only by behaviors that are
applicable to current situations. Simulation predicts the navigation performance
on a microrover in simplified Mars-analog terrain.
18: Intelligent Control and Evolution of Mobile Robot Behavior
The objective of this chapter is to describe mobile robot control systems
that integrate fuzzy reasoning with behavior control for intelligent
sensor-based navigation. A hierarchical fuzzy control architecture is
introduced that is useful for realizing complex fuzzy controllers and,
in particular, complex fuzzy behavior-based systems. Issues of behavior
synthesis, and design via the genetic programming approach to computational
evolution are discussed. Finally, applications that demonstrate the
practicality of this integration are covered which include tethered
semi-autonomous control and autonomous embedded control.
17:Introduction to Fuzzy Logic With Application
to Mobile Robotics
A brief introduction to fuzzy set theory and its application to control
systems is provided. Fuzzy sets do not have sharp boundaries and are
therefore able to represent linguistic terms which may be considered
"gray" or vague. Aspects of fuzzy set theory and fuzzy logic are highlighted
in order to illustrate distinct advantages, as contrasted to classical
sets and logic, for use in control systems. Using a mobile robot navigation
problem as an example, the synthesis of a fuzzy control system is examined.
16:Soft Computing Paradigms for Learning Fuzzy
Controllers with Applications to Robotics
Three soft computing paradigms for automated learning in robotic systems are
briefly described. The first employs Genetic Programming (GP) to evolve
rules for fuzzy-behaviors to be used in mobile robot control. The second
paradigm develops a two-level hierarchical fuzzy control structure for
flexible manipulators. It incorporates Genetic Algorithms (GA) in a
learning scheme to adapt to various environmental conditions. The third
paradigm concentrates on a methodology that uses a Neural Network (NN) to
adapt a fuzzy logic controller (FLC) in manipulator control tasks.
Simulation results of fuzzy controllers learned with the aid of these
soft computing paradigms are presented.
15:Genetic Programming of Fuzzy Coordination
Behaviors for Mobile Robots
Intelligent robot navigation can be achieved using a control system comprised
of a collection of special-purpose motion routines, or behaviors. An approach
to behavior coordination in multi-behavior systems is described with emphasis
on evolution of fuzzy coordination rules using the genetic programming (GP)
paradigm. Both conventional GP and steady-state GP are applied to evolve a
fuzzy-behavior for sensor-based goal-seeking to be used in a hierarchical
fuzzy navigation controller. The usefulness of GP is demonstrated by
simulating performance of evolved coordination rules for autonomous navigation.
14:Mobile Robot Autonomy via Hierarchical Fuzzy
Behavior Control
Realization of autonomous behavior in mobile robots, using fuzzy logic control,
requires formulation of rules which are collectively responsible for necessary
levels of intelligence. This collection of rules can be conveniently decomposed
and efficiently implemented as a hierarchy of fuzzy-behaviors. This paper
describes how this can be done using a behavior-based architecture. The
approach is motivated by ethological models which suggest hierarchical
organizations of behavior. A behavior hierarchy and mechanisms of control
decision-making are described. Simulation predicts performance and reveals
characteristics of behavior interaction.
13:On Genetic Programming of Fuzzy Rule-based Systems for Intelligent Control
Fuzzy logic and evolutionary computation have proven to be convenient tools for
handling real-world uncertainty and designing control systems, respectively. An
approach is presented that combines attributes of these paradigms for the purpose of
developing intelligent control systems. The potential of the genetic programming
paradigm (GP) for learning rules for use in fuzzy logic controllers is evaluated by
focussing on the problem of discovering a controller for mobile robot path tracking.
Performance results of incomplete rule-bases compare favorably to those of a complete
FLC designed by the usual trial-and-error approach. A constrained syntactic
representation
supported by structure-preserving genetic operators is also introduced.
12:Coordination of Distributed Fuzzy Behaviors in Mobile Robot Control
Mobile robot navigation can be achieved using a control system comprised of a
collection of special-purpose motion routines, or behaviors. In order to exhibit
robust autonomous performance, a suitable strategy for coordinating the behaviors
must be adopted. This paper describes an approach to behavior coordination and
conflict resolution within the context of a hierarchical architecture of fuzzy
behaviors. This architecture can be viewed as a network of distributed intelligent
behaviors in which coordination is achieved using weighted decision-making based on
behavioral degrees of applicability. Aspects of the coordination and conflict
resolution procedures that are based on generalized concepts from fuzzy control are
discussed. The strategy is appropriate for fuzzy control of systems that can be
represented by hierarchical or decentralized structures.
11:Hybrid Fuzzy Control Schemes for Robotic Systems
Three hybrid fuzzy control schemes for robotics applications are described. The first
scheme concentrates on a control architecture which incorporates fuzzy logic theory
into the framework of behavior control for mobile robot navigation. The second scheme
develops a two-level hierarchical fuzzy control structure for flexible manipulators.
It incorporates Genetic Algorithms (GA) in a learning scheme to adapt to various
environmental conditions. The third scheme concentrates on a methodology that uses
a Neural Network (NN) to adapt a fuzzy logic controller (FLC) in manipulator trajectory
following tasks.
10:Fuzzy Spatial Map Representation for Mobile Robot Navigation
The form of map representation employed by autonomous mobile robot navigation systems
is an important issue as it directly affects real-time map-based navigation. A method
is proposed for representing uncertain spatial information in maps used to guide
autonomous motion in real world environments. The method is based on the use of fuzzy
relations. It results in a concise representation that can be expressed in linguistic
terminology and is easy to retrieve and maintain. A simulated navigation example
demonstrates the applicability of the representation in a simple cluttered environment.
9:Fuzzy Relational Representation of Uncertain Spatial Maps for Autonomous Vehicles
Intelligent control of vehicles intended for operation in unstructured environments
is a topic of considerable interest in the area of autonomous systems. Autonomous
vehicle navigation must be achieved in the presence of the uncertainty of the real
world and the sensors used to observe it. In this paper, a method is proposed for
representing uncertain spatial information in maps used to guide autonomous vehicle
motion in real world environments. The method is based on the use of fuzzy relations
and results in a concise representation amenable to real-time autonomous navigation.
8:On Embedded Fuzzy Controllers
In recent years various practical applications of fuzzy logic control have appeared
in the literature. Many of the applications have been implemented as software
controllers utilizing a personal computer in the loop. Such control implementations
are suitable for a number of applications but are not very feasible for embedded
systems, e.g. untethered mobile robots. In this paper, we explore feasible
approaches to embedded fuzzy logic control using commercially available microcontrollers
and state-of-the-art IC chips. Some potential applications related to the authors'
work on mobile robot navigation and control are discussed. However, the fuzzy
control approaches are generally applicable to other embedded systems.
7:Fuzzy Logic and Behavior Control Strategy for Autonomous Mobile Robot Mapping
In support of research efforts to develop reactive mobile robots capable of
autonomous navigation, we are concentrating on providing human reasoning
capabilities
as a resource for intelligence and integrating the world mapping and
navigation tasks. The underlying philosophy is that an autonomous mobile robot
should be able to make a map of its environment as it navigates through that
environment. Fuzzy logic control and behavior-based navigation are proposed as
means to implement this philosophy. Fuzzy logic provides the approximate reasoning
necessary for handling the uncertainty inherent in mobile robot navigation. By
decomposing the problem into simpler tasks we can develop mobile robots that exhibit
a sufficient level of intelligence for performing navigation and mapping. In this
paper, we discuss a strategy for incorporating fuzzy reasoning into the framework
of reactive behavior control systems for autonomous mapping.
6:Manipulator Control for Rover Planetary Exploration
An anticipated goal of Mars surface exploration missions will be to survey and
sample surface rock formations which appear scientifically interesting. In such
a mission, a planetary rover would navigate close to a selected sampling site
and the remote operator would use a manipulator mounted on the rover to perform
a sampling operation. Techniques for accomplishing the necessary manipulation
for the sampling components of such a mission have been developed and are
presented. We discuss the implementation of a system for controlling a seven
(7) degree of freedom Puma manipulator, equipped with a special rock gripper
mounted on a planetary rover prototype, intended for the purpose of performing
the sampling operation. Control is achieved by remote teleoperation. This
paper discusses the real-time force control and supervisory control aspects of
the rover manipulation system. Integration of the Puma manipulator with the
existing distributed computer architecture is also discussed. The work
described is a contribution towards achieving the coordinated manipulation and
mobility necessary for a Mars sample acquisition and return scenario.
5: Use of Symbolic Computation in Robotics Education
This paper describes an application of symbolic computation in robotics
education. A software package is presented which combines generality,
user interaction, and user-friendliness with the systematic usage of
symbolic computation and artificial intelligence techniques. The software
utilizes MACSYMA, a LISP-based symbolic algebra language, to generate
automatically closed-form expressions representing forward and inverse
kinematics solutions, the Jacobina transformation matrices, robot pose
error compensation models equations, and Lagrange dynamics formulation
for N degree-of-freedom, open chain robotic manipulators. The goal of
such a package is to aid faculty and students in the robotics course
by removing burdensome tasks of mathematical manipulations. The software
package has been tested for its accuracy using commercially available
robots.
4: Computer Generation of Geometrical Error Equations Applicable for Improvement of Robots' Positioning Accuracy
A symbolic manipulation software package has been developed to automatically
generate geometrical error model equations applicable for robot's error
compensation and calibration. The software package named AREEM (Automatic
Robot Error Equation Modeler) utilizes MACSYMA, a LISP-based artificial
intelligence language, to output scalar algebraic equations representing
the positioning error correction in world coordinates for N degree-of-freedom
robots. The motivation of this work is trifold: (1) to demonstrate the
feasibility of utilizing a well-established algebraic manipulation code
for robot error compensation, simulation and modeling purposes, (2) to
provide a base for investigating the performance and accuracy of numerous
robot geometrical error compensation models; and (3) to represent output
results in a concise form eliminating completely the manual derivation
process. When such computer generated outputs are fed to a simulation
program, saving in computational time for error estimation is realized.
At present, AREEM incorporates three kinematic error models based on the
Denavit-Hartenberg representation (DH) and a non-DH representation. The
AREEM program is user-friendly, interactively menu-driven, and has been
tested on numerous robots. The worst case takes 174.80 seconds to generate
error model equations in world coordinates for the PUMA 600 robot, which
is insignificant compared to the time required for accurate manual derivation.
3: Application of Symbolic Computation in Robot Pose Error Modelling
This paper presents an application of symbolic computation in geometrical
error modeling and simulation of an industrial robotic manipulator.
A program called SCRPE
(Symbolic Computation of Robot Pose Errors) has been developed to
automatically generate error model equations for the end-effector of N
degree-of-freedom robots. The SCRPE utilizes symbolic manipulation
capability of MACSYMA (a LISP-based artificial intelligence code and
the trademark of Symbolics, Inc.). The prime reasons for this work are to
provide a base for comparison, performance assessment and accuracy judgement
of numerous error calibration and compensation models available
in the literature; and to represent output results in a concise form
eliminating completely the manula derivation process. When
such computer generated outputs are fed to a simulation program,
saving in computational time for error estimation is realized.
As an example, the mathematical error model considered here is based on
small perturbation in link parameters defined in accordance with a
classical Denavit-Hartenberg notation. The time required to compute
first and second order terms of the model are compared for numerous
robots. The worst case scenario takes 290 seconds for PUMA 600 series
robot, which is insignificant compared to the time required for accurate
manual derivation. The SCRPE program is user-friendly, interactively
menu-driven and has been developed on VAX 750 Digital computer under the
VMS operating system. Interested readers can obtain the program copy
by contacting the first author.
2: An Interactive Program for Symbolic Kinematic Solution of Industrial Robots
The mathematical derivation of kinematic equations for industrial robots
involves many matrix manipulations and results in complex expressions. When
done manually, the derivation is algebraically laborious and error-prone.
In this paper, an interactive software package is presented which eliminates
such difficulties. Symbolic Manipulation of Industrial Robot Kinematics
(SMIRK) software package combines generality, user interaction, and
user-friendliness with the systematic usage of symbolic computation
adopting the production system approach of artificial intelligence. The
program is written in MACSYMA, a LISP-based symbolic algebra language,
that derives forward and inverse kinematic solutions; and the Jacobian
transformation matrix for an n degree-of-freedom robot. The key feature
of the SMIRK program is bifold: 1) it has been designed in a way that the
program can be used as an educational aid to faculty and students of
robotics; ans 2) it can be considered as a design and computer-aided
analysis tool to investigate different manipulator configurations as
per the design need of the robotics engineer. The SMIRK program has
been sucessfully tested for its accuracy using commercially available
industrial robots.
1: Mechanization of Manipulator Kinematic Equations via MACSYMA
The mathematical derivation of kinematic equations for industrial robots
involves many matrix manipulations and results in complex expressions. When
done manually, the derivation is algebraically laborious and error-prone.
In this paper, an interactive software package is presented which eliminates
such difficulties. Symbolic Manipulation of Industrial Robot Kinematics
(SMIRK) software package combines generality, user interaction, and
user-friendliness with the systematic usage of symbolic computation
adopting the production system approach of artificial intelligence. The
program is written in MACSYMA, a LISP-based symbolic algebra language,
that derives forward and inverse kinematic solutions; and the Jacobian
transformation matrix for an n degree-of-freedom robot. The key feature
of the SMIRK program is bifold: 1) it has been designed in a way that the
program can be used as an educational aid to faculty and students of
robotics; ans 2) it can be considered as a design and computer-aided
analysis tool to investigate different manipulator configurations as
per the design need of the robotics engineer. The SMIRK program has
been sucessfully tested for its accuracy using commercially available
industrial robots.
Back to publications list
Home
E-Mail:
tunstel@robotics.jpl.nasa.gov