DRAFT Wed. 9/29 2:02 a.m. 4. Statement of Justification

somberastonishingΤεχνίτη Νοημοσύνη και Ρομποτική

13 Νοε 2013 (πριν από 3 χρόνια και 6 μήνες)

72 εμφανίσεις

DRAFT


Wed. 9/29 2:02 a.m.


4. Statement of Justification



The work proposed herein addresses several issues critical to and unique to OExS. This work
pertains especially to several key aspects of Spiral 2, human lunar return by 2020.


The enabling tec
hnology we propose is software
-
defined radio (SDR). This is a communications
hardware/software platform in which the radio frequency (RF) signal is generated or received
directly by a digital signal processor (DSP), so that the translation to/from baseban
d is defined in
software. This provides the capability to change modulation and encoding on the fly, with
consequent capability to perform sophisticated tradeoffs impossible or prohibitively expensive
with a conventional RF front
-
end. An SDR transceiver ca
n, in principle, communicate with
virtually any other type of transceiver, so that evolving technology no longer renders older
equipment obsolete, and newer encoding and modulation techniques can be implemented without
replacing hardware.


The application
of software defined radio (SDR) to space telecommunications directly addresses
the issues of affordability and sustainability, which are identified as the first key objective in the
H&RT Formulation Plan. Software
-
defined radio also provides flexibility,
as well as reusability,
modularity, and reconfigurability, all of which are key strategic technical challenges identified in
Section 6.4 of the Formulation Plan. This is identified in the BAA in Appendix A, Section 1.3 as
being of interest to the Communica
tions, Computing, Electronics and Imaging (CCEI) Element,
specifically in the Space Communications and Networking Theme. A stated goal is " ... to
achieve sustainable, scalable, fully accessible and fully reliable and secure communications and
networking i
nfrastructure within the solar system for multiple robotic and human assets wherever
they are deployed. " Furthermore, one of the technologies specifically named is "software radio
based technologies for flexible, energy efficient, multi access application
s."


We propose to apply SDR specifically to robotic / human networks, which may contain members
physically widely separated. This results in "limited human oversight due to distance," which is a
challenge identified in the Formulation Plan as "unique to N
ASA" (Ref. Section 7.2.7). In the
BAA, this is identified in Appendix A, Section 1.4, which describes the interest on the part of the
Software, Intelligent Systems, and Modeling (SISM) Element in the areas of Autonomy and
Intelligence and Multi
-
Agent Teami
ng, specifically " multi
-
robotic teams constructing planetary
or orbital facilities."


5. Proposal Abstract


Project Description


The scope and aim of the proposed project is to develop a simulated interplanetary exploration
environment in which a laborat
ory
-
based heterogeneous colony of cooperative robots interact,
enabled by software radio. Communication latency will be introduced artificially to mimic time
delays consistent with interplanetary missions. Members of the colony will be use a variety of
co
mmunication modes in order to simulate a realistic scenario useful to NASA in which
communicating entities most likely will not be equipped with identical transceivers.


We propose to build on our current expertise in the areas of software radio and cooper
ative
robotics to develop a cooperative autonomous assembly and exploration system in support of
NASA exploration goals in the H&RT program. It is envisioned that the communication
infrastructure based on software radio will enable "plug and play" function
ality, so that insertion
or deletion of specific communicating entities into the communication scheme will be
transparent.


The specific H&RT Formulation (Section 6.4.2) Strategic Technical Challenges addressed by
this proposal are:

1. Robotic Networks

2.

Modularity

3. Autonomy

4. Margins and Redundancy

5. Data
-
Rich Virtual Presence


Project Goals:

Goal 1: Establish an interactive network (colony) of at least three and potentially as many as
twenty robots equipped with software radio communications capabi
lity

Goal 2: Identify key features and protocol necessary to mimic the interplanetary exploration
environment from a communications standpoint as realistically as possible.

Goal 3: Use the testbed to develop and refine algorithms necessary to eventually ac
complish
specific tasks, such as assembly of mechanical parts, spacecraft docking, surface exploration, and
others as suggested by NASA sponsors.


H&RT Goals and Objectives Supported:

1. Space Backbone Networks and Space Wide Area Networks (Communications
, Computing,
Electronics, and Imaging
-

CCEI). This project will enable the interaction of numerous
autonomous entities over both short distances and interplanetary distances.

2. Multi
-
Agent Teaming (Software, Intelligent Systems, and Modeling
-

SISM). Th
is project will
enable multiple entities to cooperate on tasks over distances ranging from short range (a few
meters) to interplanetary.


Technology Maturation Approach, Challenges, and Teaming


This project will begin at a TRL of 2 (Concept formulation) a
nd end at TRL 4 (Laboratory
breadboard). The development of the proposed robot colony will begin with a paper study to
identify the key components required in consultation with the team members and NASA
sponsors. Key expected obstacles include:

(1) Identif
ying the operating frequency band(s) and modulation methods most suited to the target
application

(2) Implementing the necessary electronics in the limited size, weight, and cost constraints
necessary for the laboratory testbed environment.

(3) Adapting ex
isting protocols to accommodate latencies varying from microseconds to hours
due to the need to interact over distances ranging from a few meters to interplanetary distances.


Technology Maturation:

The most likely road to technology maturation is through
the Advanced Space Operations
Technology (ASO) program element, although there is considerable potential through the Lunar
and Planetary Surface Operations (LPSO) program element as well. In particular, the proposed
research can contribute to ASO in the ar
eas of in
-
space assembly, autonomy, reconfigurability,
and data
-
rich virtual presence, and can contribute to LPSO in the areas of intelligent and agile
surface mobility systems, surface manufacturing and construction systems, and surface
environmental mana
gement.


Teaming:

The project will be directed by Dr. Thaddeus Roppel (Lead, Auburn University Sensor Fusion
Laboratory). Dr. Roppel and Dr. Agrawal (Co
-
Lead, Wireless Research Center, Auburn
University) will be primarily involved with software radio imple
mentation, including firmware
and software development. Dr. Wilson (Co
-
Lead, U. of Washington) will contribute expertise
from the area of distributed architecture implementation, both software and hardware, as well as
robotic systems. Dr. Bradley (Co
-
Lead,

NASA LaRC) will contribute in the area of robotics,
telerobotics, NASA goals, and NASA policies and procedures.


Impact on Future Exploration Systems


Exploration systems of the future will undoubtedly involve multiple vehicles fanning out over
planet sur
faces, together with numerous manned and unmanned entities in orbit or in transit
between planets. In any conceivable scenario, it will be of the utmost importance for each entity
to have the ability to communicate with all or a subset of the others with h
igh reliability.
Furthermore, tasks such as exploration, assembly, and inspection will need to be accomplished
through cooperation and will require highly flexible "plug and play" communication systems as
proposed here.

6.
Project Description



a.

Desc
ription of the proposed technology project including specific goals and objectives by
phase and contract period.


The proposed project is the development of a simulated interplanetary exploration environment
in which a laboratory
-
based heterogeneous colony

of cooperative robots interact, enabled by
software
-
defined radio (SDR). Communication latency will be introduced artificially to mimic
time delays consistent with interplanetary missions. Members of the colony will use a variety of
communication modes i
n order to simulate a realistic set of scenarios useful to NASA.




Relationship between the proposed work and the research emphases defined in the BAA
(Appendix A)


$COMPLETE TABLE HERE


PROPOSED TASKS

BAA Research Emphases

Relationship









Phase 1
:

In Phase 1 (12 months), we intend to design and build at least 4 prototype SDR boards (allowing
for some design flexibility at the hardware level) using COTS components and install them on
one fixed and two mobile robots or small rovers, as well as at
a master station with human
interface. We plan to implement a relatively simple digital modulation scheme, most likely
$JAEGER at low bit rate $JAEGER. We will conduct several types of experiments and
demonstrations, including variable
-
delay control loops,

intelligent real
-
time adjustment of
communication parameters, and translation
-

using an SDR to bridge a link between two other
colony members using different protocols.


Phase 1 Goals:

1. Demonstrate, through a small
-
scale effort, the benefits of softwar
e
-
defined radio applied to
human and robotic networks.

2. Demonstrate the capability of our team to organize and manage the proposed work effectively,
and to deliver the promised work product.

3. Lay the groundwork for the Phase 2 effort.


Phase 1 Objectiv
es:

1. Have in place an operational four
-
node human / robotic network communicating with
software
-
defined radio.

2. Have in place a working demonstration of two simple, NASA
-
relevant cooperative activities
within the colony which could not be achieved eff
ectively without employing SDR: (i)
intelligent real
-
time adjustment of communication parameters to overcome channel degradation,
and (ii) translation
-

using an SDR to bridge a link between two other colony members using
different protocols.

3. Have a des
ign for new DSP boards for Phase 2, Year 1 essentially finished.



Phase 2:

In Phase 2 (36 months), we intend to enlarge the human
-
robotic network from Phase 1 to include
at least 2 human
-
interface nodes, 2 stationary robotic nodes, and 3 mobile robotic no
des (the
latter to be delivered by team member Langley Research Center). This will provide a reasonable
level of complexity to investigate the scenarios of interest, while not being excessively expensive
to construct or maintain. Improved SDR boards desig
ned during Phase 1 will be fabricated and
installed at each node. We will establish a design cycle for SDR boards to take advantage of new
DSP technology as it becomes available. We anticipate obtaining new DSP's near the start of
each contract year, and r
edesigning the SDR to take advantage of their increased throughput.


The complexity of the tasks to be demonstrated will increase each year in Phase 2. A final year
task which can serve as a point of discussion would be cooperative assembly of a structure
under
remote (large time delay), minimal human guidance.


Phase 2 Goals:

1. Demonstrate the benefits and optimal usage of software
-
defined radio applied to human and
robotic networks.

2. Establish a permanent testbed for experiments in software defined rad
io


enabled cooperative
robotics which would be available for the benefit of NASA and the general use of the scientific
and technical community.


Phase 2 Objectives:

1. Have in place an operational seven
-
node human / robotic network communicating with
so
ftware
-
defined radio.

2. Have in place a working demonstration of cooperative assembly
of a structure under remote
(large time delay), minimal human guidance.

3.
Have in place a working demonstration of a servicing operation in which a rover on a planet
su
rface is to be serviced by a newer, SDR
-
equipped machine. The scenario requires the two to
communicate, and requires some remote human intervention.


b. Description of the technology development/maturation approach, including, specific
technical challenges

that you expect to encounter, and technology metrics.


The most likely road to technology maturation is through the Advanced Space Operations
Technology (ASO) program element, although there is considerable potential through the Lunar
and Planetary Surfac
e Operations (LPSO) program element as well. In particular, the proposed
research can contribute to ASO in the areas of in
-
space assembly, autonomy, reconfigurability,
and data
-
rich virtual presence, and can contribute to LPSO in the areas of intelligent a
nd agile
surface mobility systems, surface manufacturing and construction systems, and surface
environmental management.


SDR Technical Challenges:



For SDR, the primary challenge is to obtain the fastest possible analog
-
to
-
digital (A/D)
conversion to direc
tly convert the received RF signal to baseband. Since we intend to use
off
-
the
-
shelf DSP's, we shall depend to a great extent upon "Moore's Law" (the observed
doubling of many technology metrics approximately every 18 months) to yield faster
DSP's.



A rela
ted SDR design technical challenge is to manage the large number of samples that
result from the direct down
-
conversion, especially for more complex modulation and
encoding schemes. This is typically addressed in SDR design by employing FPGA's or
ASIC's fo
r improved throughput.

For Cooperative Robotics, we identify three main technical challenges:



How is the timing of each node's actions related to each other node (synchronization)?



How is latency to be handled?



How can the internal behavior of the networ
k be understood in a way that allows the
designer to intelligently make improvements?


Technology Metrics: Acceptance criteria for assessing progress/accomplishment for key
milestones


KEY MILESTONES

Approx.
Contract
Month

Abbrev.
on Gantt
Chart

Technology

Metrics

Phase 1




2
-
node human / robotic
network communicating with
software
-
defined radio.


6

2N
-
10

Communication bit rate
of 10 kbits second

4
-
node human / robotic
network communicating with
software
-
defined radio.


Demonstrate translation
-

using an

SDR
-
equipped robot
(R2) to bridge a link between
two other colony members
(R1 and R3) which have
different communication
protocols from each other.


12

4N
-
50




Demo

Trans

Communication bit rate
of 50 kbits second



R1 and R3 enabled to
communicate at an
effective bit rate of 10
kbits/sec.


Phase 2




Cycle 2 SDR boards designed

18

Design
-
2

100 kbits/s

Three (3) LaRC Mobile
robots operational

24

3MR
-
50

Capable of transceiving
using SDR at 50 kbits/sec
and fully controlled
motion

4
-
node human / robotic
network communicating with
software
-
defined radio.


Cycle 3 SDR boards designed

30

4N
-
100




Design
-
3

Communication bit rate
of 100 kbits second



200 kbits w/ advanced
modulation and encoding

Demonstration of robotic
cooperative assembly
of a
structure w
ith human
guidance. Human selects
structure, two robots assemble
it, third robot brings parts.

36

Demo
Assem

Structure consisting of
ten 30 cm long, 1 cm
diam. alum. rods
constructed in 1 hour.

Demonstration of a servicing
operation in which a rover on
a
planet surface (R1) is to be
serviced by a newer, SDR
-
equipped machine (R2). The
scenario requires the two to
communicate, and requires
some remote human (H)
intervention.


42

Demo
Service

Operation completed
successfully 4 out of 5
tries.

7
-
node human /
robotic
network communicating with
software
-
defined radio.

48

7N
-
200

Communication bit rate
of 200 kbits second



c. A description of the impact of the proposed technology to future exploration systems,
including specific benefits to exploration, and over
all long
-
term use (e.g. future systems live
Crew Exploration Vehicles, lunar rovers, lunar
-
planetary bases, etc)


Exploration systems of the future will undoubtedly involve multiple vehicles fanning out over
planet surfaces, together with numerous manned a
nd unmanned entities in orbit or in transit
between planets. In any conceivable scenario, it will be of the utmost importance for each entity
to have the ability to communicate with all or a subset of the others with high reliability.
Furthermore, tasks s
uch as exploration, assembly, and inspection will need to be accomplished
through cooperation and will require highly flexible "plug and play" communication systems as
proposed here. The proposed work provides a demonstrative, prototype infrastructure for

sustainable missions of the type described in the Formulation Plan. It also provides for
reconfigurability and adaptability to new environments and unforseen circumstances.


7. Statement of Work (SOW)

This section provides a Statement of Work (SOW) segreg
ated by Phase 1 and Phase 2 (yearly).
The following sections are included: (7.1) Scope (7.2) Objectives (7.3) SOW tasks organized in
a Work Breakdown Structure (WBS), (7.4) Program Schedule & Milestones, (7.5) Acceptance
criteria (e.g. key technology me
trics) for assessing progress/accomplishment for key milestones,
and (7.6) deliverables, which shall are defined and described under the applicable task/WBS
portion of the SOW.


7.1 Scope


The scope and aim of the proposed project is to develop a simulate
d interplanetary exploration
environment in which a laboratory
-
based heterogeneous colony of cooperative robots interact,
enabled by software radio. Communication latency will be introduced artificially to mimic time
delays consistent with interplanetary
missions. Members of the colony will be use a variety of
communication modes in order to simulate a realistic scenario useful to NASA in which
communicating entities most likely will not be equipped with identical transceivers. The project
will begin at te
chnology readiness level (TRL) 2 and end at TRL 4, as defined in the BAA (ref.
Fig. 1).


7.2 Objectives


Phase 1 Objectives:

1. Have in place an operational four
-
node human / robotic network communicating with
software
-
defined radio.

2. Have in place a wo
rking demonstration of two simple, NASA
-
relevant cooperative activities
within the colony which could not be achieved effectively without employing SDR: (i)
intelligent real
-
time adjustment of communication parameters to overcome channel degradation,
and
(ii) translation
-

using an SDR to bridge a link between two other colony members using
different protocols.

3. Have a design for new DSP boards for Phase 2, Year 1 essentially finished.


Phase 2 Objectives:

1. Have in place an operational seven
-
node human

/ robotic network communicating with
software
-
defined radio.

2. Have in place a working demonstration of cooperative assembly
of a structure under remote
(large time delay), minimal human guidance.

3.
Have in place a working demonstration of a servicing o
peration in which a rover on a planet
surface is to be serviced by a newer, SDR
-
equipped machine. The scenario requires the two to
communicate, and requires some remote human intervention.


7.3 Work Breakdown Structure (WBS)


The WBS for this project is as

follows (identical for Phase 1 and Phase 2):


Level 1


Project



Level 2: Work Products A


E


A. Software
-
defined radio (SDR)

B. Cooperative telerobotics software environment (CTSE)

C. Human and robotic physical nodes (HRN)

D. System monitoring softwar
e environment (SMSE)

E. Project management, project reports and documentation (PRD)



The Level 2 Work Products A


E are described in detail in the following subsections.


7.3.1 Work Product A: Software
-
defined radio (SDR)

The software
-
defined radio work
product will consist of a printed wiring board (PWB)
containing a digital signal processor (exact model to be determined as part of investigation),
several RF components, e.g., bandpass filter, power amplifier, as determined in the design effort,
and baseb
and electronics such as buffer amplifier(s) and power management. The design will be
a collaboration between Auburn University (Dr. Roppel, Dr. Jaeger) and commercial partner
CoachComm (Mr. Turkington), with some input from LaRC (Mr. Arthur Bradley at LaRC

has
considerable RF design experience).


Task A, Phase 1

The initial design for Phase 1 will target a throughput of 10 kbits/s with a conventional
modulation scheme such as quadrature phase
-
shift keying (QPSK). This will provide for a rapid
turnaround to
permit the Phase 1 milestones and demonstrations to be achieved, and to
demonstrate proof of concept.


Task A, Phase 2

The Phase 2 effort in this area will principally be to periodically redesign the SDR to take
advantage of expected increasing throughput
off
-
the
-
shelf DSP's. Redesigns will be initiated at
approximately 18 months and 30 months into the overall contract period (6 months and 18
months into Phase 2). This re
-
design will have two thrusts: operation at higher RF frequencies
and higher bit rat
es, and increased mode complexity. The exact RF modes to be considered will
be determined as part of the investigation, but may include direct
-
sequence or frequency
-
hopping
spread spectrum (DSSS, FHSS) among many other possibilities.


An additional paper
study will develop a roadmap for applications beyond the 4
-
year term of the
proposal based on further anticipated improvements in COTS DSP hardware.


7.3.2 Work Product B: Cooperative telerobotics software environment (CTSE)


This will be primarily the res
ponsibility of team member Vanderbilt University Center for
Intelligent Systems (Kaz Kawamura, Mitch Wilkes). Auburn University will contribute by
helping to integrate the software into the robot and human nodes.


Task B, Phase 1

During Phase 1, we will i
mplement a software module that can reside on each node (human or
robot) to enable cooperative behavior. This is intended to be general in nature to enable the
implementation of a wide variety of interesting scenarios. Example scenario: Human operator
des
ires to teleoperate Robot 1, but communication hardware is incompatible. Robot 2 is
equipped with SDR that can "translate" by alternating between the human node's communication
mode and Robot 1's communication mode in real time under software control. Ther
e are time
delays TD1 from human to Robot 1 and TD2 between Robot 1 and Robot 2.


Using existing code as much as possible, we will implement software modules to facilitate
cooperative behavior and support high level human commands. Modules will reside on
each
node, both human or robot centered nodes. Where appropriate these modules will support
human robot interaction through graphical user interfaces. Initially the modules will support
teleoperation of the robot by a remote human user. In Phase 2 this
capability will be expanded.


Task B, Phase 2

We will develop a software environment with sufficient flexibility to allow a variety of
cooperation scenarios to be investigated. These may include "virtual presence" applications,
collaborative assembly of st
ructures, and coordinated sensing with heterogeneous robot sensor
platforms, among others. Additionally, as mentioned in the Phase 1 description, we will add
some ability for the robot to act with limited autonomy, yet remain under the supervision of a
re
mote human user. This mode, often called teleassistance, is very practical in the presence of
large communication delays. Such delays can make direct teleoperation very difficult and
frustrating for the user. Giving the robot limited autonomy to perform

basic simple tasks enables
the user to direct the actions of the robot in terms of these simple tasks. This type of problem
decomposition is more robust in the presence of communication delays.


Hardware needs for phase 2: Virtual reality (VR) or heads
-
up display, force feedback actuators,
cameras, thermal imaging cameras, powerful computing nodes.


We will develop a software environment with sufficient flexibility to allow a variety of
cooperation scenarios to be investigated. These may include "virtual

presence" applications,
collaborative assembly of structures, and coordinated sensing with heterogeneous robot sensor
platforms, among others. Additionally, as mentioned in the Phase 1 description, we will add
some ability for the robot to act with limit
ed autonomy, yet remain under the supervision of a
remote human user. This mode, often called teleassistance, is very practical in the presence of
large communication delays. Such delays can make direct teleoperation very difficult and
frustrating for th
e user. Giving the robot limited autonomy to perform basic simple tasks enables
the user to direct the actions of the robot in terms of these simple tasks. This type of problem
decomposition is more robust in the presence of communication delays.


Hardwa
re needs for phase 2: Virtual reality (VR) or heads
-
up display, force feedback actuators,
cameras, thermal imaging cameras, powerful computing nodes.


7.3.3 Work Product C: Human and robotic physical nodes (HRN)


The construction of the physical network n
odes (colony member hardware) will be primarily
performed at Auburn University in Phase 1, but represents a significant part of the effort by
LaRC during Phase 2.


Task C, Phase 1

Auburn University and LaRC will construct three stationary robot nodes and o
ne human
-
interface node each designed to interface with the SDR designed in Task A. The stationary robot
nodes will have, at a minimum, one gripper arm, a color video /still digital camera on a rotating
mount, and an ultrasonic range / presence sensor. The

human
-
interface node will be a computer
workstation capable of running the software developed in Tasks B and D. This workstation will
also be modified and equipped as required to interface to the SDR developed in Task A.


During Phase 1, LaRC will also de
sign the mobile robots to be constructed in Phase 2.


Task C, Phase 2


During the first year of Phase 2, team member LaRC (Robotics and Intelligent Machines
Laboratory
-

Dr. Arthur Bradley) will build three mobile robots to be used as subsequent mobile
net
work nodes equipped with SDR from Task A. During years 2 and 3, these robots will be
refined and outfitted with various sensors and actuators, which may include infrared cameras,
tactile feedback, and magnetometers among many other possibilities. The desi
gn details will
depend largely on results of ongoing investigation.


During Phase 2, Auburn University will focus on improving the stationary robot nodes
constructed in Phase 1, with the goal of making them as agile and responsive as possible.


7.3.4 Work
Product D: System monitoring software environment (SMSE)


Task D, Phase 1


Task D, Phase 2


7.3.5 Project management, project reports and documentation (PRD)


Task E, Phase 1


Task E, Phase 2