Information about http://nmp.jpl.nasa.gov/ds1/DS1_Primary_Mission.pdf
50th International Astronautical Congress, Amsterdam, The…
Tags: acta astronautica, astronautical, autonomous systems, california institute of technology, deep space 1, earth science programs, flight projects, future space, instrument packages, integrated science, jet propulsion laboratory, livesay, millennium program, new millennium, science instrument, science introduction, solar electric propulsion, technology payload, technology validation mission, varghese,Pages: 13
Language: english
Created: Fri Sep 1 15:03:50 2000
Display cached document
Page 1
Page 2
Page 3
Page 4
Page 5
Page 6
Page 7
Page 8
Page 9
Page 10
Page 11
Page 12
Page 13
50th International Astronautical Congress, Amsterdam, The Netherlands, 4-8 October, 1999, IAA-99-IAA.11.2.01
Acta Astronautica 47, p. 475 (2000).
RESULTS FROM THE DEEP SPACE 1 TECHNOLOGY VALIDATION MISSION
Marc D. Rayman, Philip Varghese, David H. Lehman, and Leslie L. Livesay
Jet Propulsion Laboratory
California Institute of Technology
4800 Oak Grove Dr.
Pasadena, CA 91109 USA
Launched on October 24, 1998, Deep Space 1 (DS1) is the first mission of NASA's New
Millennium program, chartered to validate in space high-risk, new technologies important for future
space and Earth science programs. The advanced technology payload that was tested on DS1 comprises
solar electric propulsion, solar concentrator arrays, autonomous on-board navigation and other
autonomous systems, several telecommunications and microelectronics devices, and two low-mass
integrated science instrument packages. The technologies were rigorously exercised so that subsequent
flight projects would not have to incur the cost and risk of being the first users of these new capabilities.
The performances of the technologies are described as are the general execution of the mission and plans
for future operations, including a possible extended mission that would be devoted to science.
INTRODUCTION missions by developing and validating some of
the high-risk, high-benefit technologies they need.
NASA's plans for its space and Earth NMP conducts deep space and Earth orbiting
science programs call for many scientifically missions focused on the validation of these
compelling, exciting missions. To make such technologies. The spacecraft flown by NMP are
programs affordable, it is anticipated that small not intended to be fully representative of the
spacecraft, launched on low-cost launch vehicles spacecraft to be used in future missions, but the
and with highly focused objectives, will be used advanced technologies they incorporate are. As
for many of the missions. To prevent the loss of each NMP mission is undertaken, the risk of
capability that may be expected in making using the technologies that form its payload
spacecraft smaller and developing and operating should be substantially reduced because of the
missions less expensively, the introduction of knowledge gained in the incorporation of the new
new technologies is essential. capability into the spacecraft, ground segment,
and mission design as well as, of course, the
With many spacecraft carrying out its quantification of the performance during the
programs of scientific exploration, NASA could flight.
accept a higher risk per mission; the loss of any
one spacecraft would represent a relatively small Although the objective of NMP technology
loss to the program. Nevertheless, the use of new missions is to enable future science missions,
technologies in space science missions forces the NMP missions themselves are not driven by
first users to incur higher costs and risks. The science requirements. They are dedicated to
concomitant diversion of project resources from technology, with the principal requirements
the scientific objectives of the missions can be coming from the needs of the advanced tech-
avoided by certification of the technologies in a nologies they are testing. The science return from
separate effort. NMP missions is in the subsequent science
missions that become feasible.
Overview of New Millennium
The New Millennium program (NMP) is By their very nature, NMP projects are high
designed to accelerate the realization of ambitious risk. The key technologies that form the basis for
each mission are the ones which require validation
to reduce the risk of future missions. Indeed, if
Copyright 1999 by the American Institute of Aeronautics an advanced technology does not pose a high risk,
and Astronautics, Inc. The U.S. Government has a testing by NMP is not required. In many cases,
royalty-free license to exercise all rights under the these unproven technologies will not have func-
copyright claimed herein for Governmental purposes. All tionally equivalent back-ups on their test flights.
other rights are reserved by the copyright owner. Nevertheless, the failure of a new technology on
1
an NMP mission, even if it leads to the loss of the The first criterion clearly indicates that the
spacecraft, does not necessarily mean the mission goal of the mission is to determine how well the
is a failure. If the nature of the problem with the technologies work. Indeed, the wording reflects
technology can be diagnosed, the goal of pre- the recognition of the high risk of the technologies
venting future missions from accommodating the by allowing for the possibility that some might not
risk can be realized. Showing that a technology is be operable.
not appropriate for use on subsequent science
missions would be a very valuable result of an A twelfth technology, a miniature integrated
NMP flight. The acquisition of this information ion and electron spectrometer, was not included in
would achieve the goal of reducing the cost and the success criteria, because it was so late in being
risk to candidate future users of the technology. delivered that even six weeks before launch it was
Of course, it is likely that such a determination uncertain whether the device would be ready.
would lead to modifications of the implementation (This is another facet of the risk in planning to fly
of the technology, thus restoring its potential with advanced technologies.) Nevertheless, it
value to future space science missions. was delivered and has performed very well.
Overview of DS1 All the technologies except autonomous
Deep Space 1 (DS1) is the first project of navigation received 100% or more of their
NMP. Its payload consists of 12 technologies. required testing by the end of June 1999. An
The criteria for "complete mission success," asteroid encounter planned for July 29, 1999 tests
agreed to by NASA Headquarters and JPL, are: 5% of the autonomous navigation system.
1) Demonstrate the in-space flight operations and In addition to its technical objectives, DS1
quantify the performance of the following 5 was intended to probe the limits of rapid develop-
advanced technologies: ment for deep-space missions. The initial study
- Solar electric propulsion (SEP) of DS1 was undertaken only 39 months before
- Solar concentrator arrays launch, an unprecedentedly short time for a
- Autonomous navigation NASA deep-space mission in the modern era. At
- Miniature camera and imaging spectrometer the time the preliminary concept study was
- Small deep space transponder initiated, the only definition of the project was that
it should validate solar electric propulsion and
and any 3 of the following 6 advanced other unidentified technologies in deep space and
technologies: that launch should occur sometime in 1998. The
- Ka-band solid state power amplifier level-1 requirements and goals1 were formulated
- Beacon monitor operations 26 months prior to launch.
- Autonomous remote agent
- Low power electronics Further background on the project,
- Power actuation and switching module including the selection of technologies and the
- Multifunctional structure mission and spacecraft design, and additional
information on NMP are presented elsewhere.1,2
2) Acquire the data necessary to quantify the
performance of these advanced technologies by TECHNOLOGY RESULTS
September 30, 1999. Analyze these data and
disseminate the results to interested Overviews of DS1's advanced technologies
organizations/parties by March 1, 2000. and the results from flight testing follow. The
mission in which the technologies were used is
3) Utilize the on-board ion propulsion system discussed in the next section.
(IPS) to propel the DS1 spacecraft on a trajectory
that will encounter an asteroid in fiscal year 1999. Solar electric propulsion
Solar electric propulsion (SEP) offers sig-
4) Assess the interaction of the IPS operations nificant mass savings for future deep-space and
with the spacecraft and its potential impact on Earth-orbiting spacecraft that require substantial
charged particle, radio waves and plasma, and velocity changes. The objective of the NSTAR
other science investigations on future SEP- (NASA SEP Technology Application Readiness)
propelled deep space missions. program, to validate low-power ion propulsion,
2
was a good match to NMP's goals. NSTAR The IPS operated over a broad range of its
involved a collaboration among JPL, NASA's 112 throttle levels, from input powers of 580 W
Glenn Research Center, Hughes Electron (throttle level 6) to 2140 W (throttle level 90). The
Dynamics, Spectrum Astro, Moog, and Physical corresponding specific impulses were 1975 s and
Science, Inc. 3180 s. Measured thrust (determined through
radio navigation) was within 2% of the prelaunch
The ion propulsion system (IPS) on DS13 prediction throughout the range.
uses a hollow cathode to produce electrons to
collisionally ionize xenon. The Xe+ is electro- Comparison with the extensive ground-test
statically accelerated through a potential of up to program showed that operation in space is more
1280 V and emitted from the 30-cm thruster benign and contamination is lower. The vast
through a pair of molybdenum grids. A separate body of data from the diagnostics sensors on the
electron beam is emitted to produce a neutral effects of the IPS allows the development of
plasma beam. The power processing unit (PPU) guidelines for future designers on how to make
of the IPS can accept as much as 2.5 kW, corre- fields and particles measurements on future IPS-
sponding to a peak thruster operating power of propelled spacecraft without interference from the
2.3 kW and a thrust of 92 mN. Throttling is propulsion system.
achieved by balancing thruster electrical
parameters and Xe feed system parameters at In the first attempt to thrust with the IPS (on
lower power levels; and at the lowest PPU input November 10, 1998), it operated for about 4.5
power, 525 W, the thrust is 19 mN. The specific minutes and then switched to a standby mode. It
impulse ranges from 3200 s with about 2 kW is believed that the unplanned termination of the
delivered to the PPU to 1900 s at the minimum thrusting was the result of a contaminant causing a
throttle level. short between the two grids. Attempts to restart
the thruster on that day were unsuccessful.
Because the purpose of flying the IPS was to Thermal cycling during the subsequent two weeks
validate it for future space science missions, a changed the spacing between the grids, thus
comprehensive diagnostic system is also on the stressing the contaminant, and when the IPS was
spacecraft.4 This aided in quantifying the inter- commanded on again it operated as desired.
actions of the IPS with the spacecraft, including Similar phenomena have been observed with other
advanced-technology science instruments, and ion thrusters in space.
validating models of those interactions. The
diagnostic instrument suite includes a retarding In the 1799.4 hours of thrusting (for
potential analyzer, two Langmuir probes, search- deterministic thrust, trajectory correction
coil and fluxgate magnetometers, a plasma wave maneuvers, and dedicated tests), the total Xe
sensor, and two pairs of quartz-crystal micro- consumption was 11.4 kg, providing 699.6 m/s.
balances and calorimeters. One of these pairs has After the first day of unsuccessful attempts to
a direct view of the ion thruster exit, while the resume thrusting, all 34 IPS starts in the mission
other is shadowed by spacecraft structure. were successful.
Measurements included the rate and extent of
contamination around the spacecraft from the Xe+ All spacecraft systems operated normally
plume and the sputtered Mo from the grid, electric during IPS thrusting. Telecommunications during
and magnetic fields, and the density and energy of IPS thrusting, even with the radio signal passing
electrons and ions in the vicinity of the spacecraft. through the plasma, were unaffected. Sensors
As a bonus, the sensors will be used to comple- 0.7 m from the thruster with a direct line of sight
ment science measurements of DS1's ion and to the exit grid recorded about 10 nm of surface
electron spectrometer (see below) during the small contamination. Nearby sensors, without a direct
body encounters. line of sight, accumulated an order of magnitude
less.
By June 30, 1999, the IPS had operated for
nearly 1800 hours. This included several Solar concentrator array
dedicated tests, but the majority of the time was Because of the IPS, DS1 required a high-
devoted to placing the spacecraft on a trajectory to power solar array. The Ballistic Missile Defense
reach asteroid 1992 KD (in accordance with the Organization (BMDO), working with NASA's
third mission success criterion). Glenn Research Center, AEC-Able Engineering,
3
and Entech, developed the Solar Concentrator system validated on DS1, AutoNav6, began
Array with Refractive Linear Element Technology functioning immediately upon activation of the
(SCARLET II).5 BMDO wanted a flight test for spacecraft after separation from the launch
SCARLET, and because it could provide the vehicle, which occurred in Earth's shadow. The
necessary high power, including it on DS1 was attitude control system (ACS) used a commercial
mutually beneficial. star tracker to determine its attitude. Then the
real-time part of AutoNav correctly provided ACS
SCARLET uses cylindrical silicone Fresnel with the position of the Sun so that ACS could
lenses to concentrate sunlight onto GaInP2/GaAs/ turn the spacecraft to the attitude needed to
Ge cells arranged in strips. Including the optical illuminate the solar arrays upon exiting the
efficiency of the lenses, a total effective magnifi- shadow.
cation greater than 7 is achieved. With relatively
small panel area actually covered by solar cells, the Data stored on board for use by AutoNav
total cost of cells is lowered, and thicker cover include a baseline trajectory, generated and
glass becomes practical, thus reducing the optimized on the ground; the ephemerides of the
susceptibility to radiation. The dual junction cells DS1 target bodies, distant "beacon" asteroids,
display significant quantum efficiencies from 400 and all planets except Pluto; and a catalog of the
nm to 850 nm, and achieved an average efficiency positions of 250,000 stars (all contained in the
in flight of about 22.5%. Tycho catalog).
The pair of arrays produced 2.5 kW at 1 AU, Throughout the mission, about once per
within 1% of the prelaunch prediction. Each array week, AutoNav was invoked by the operating
comprises four panels that were folded for launch, sequence to allow it to acquire optical navigation
and a single-axis gimbal controls pointing in the images. It issues commands to ACS and the
more sensitive longitudinal axis. The two wings integrated camera and imaging spectrometer (see
include a total of 720 lenses, each focusing light below) to acquire visible-channel images, each
onto 5 cells. DS1 is the first spacecraft to rely with one beacon asteroid and known background
exclusively on refractive concentrator arrays; it stars. On-board image processing allows accurate
also is among the first to use only multibandgap extraction of the apparent position of each asteroid
cells. with respect to the stars, thus allowing the space-
craft location to be estimated. The heliocentric
The array is one of the three new tech- orbit is computed with a sequence of these
nologies that had to work correctly immediately position determinations combined with estimated
after launch in order for the mission to proceed; solar pressure, calculated gravitational pertur-
stored battery energy was sufficient only for a few bations, and on-board knowledge of the thrust
hours. A substantial part of the validation of the history of the IPS and incidental accelerations
array was the mechanical deployment and subse- from unbalanced turns by the hydrazine-based
quent pointing. The deployment was so accurate reaction control system (RCS). The trajectory
that, following dedicated tests, no pointing then is propagated to the next encounter target, and
adjustments were deemed necessary, and the array course changes are generated by the maneuver
provided stable operation throughout the mission. design element. In general, those course correc-
tions are implemented through changes in the IPS
Autonomous navigation thrust direction and duration, but in certain cases
Because mission operations is a significant described below, the maneuvers are accomplished
part of its science budget, NASA explicitly with dedicated IPS or RCS maneuvers.
included autonomy in its guidelines to NMP. A
reduction in requirements for Deep Space After AutoNav parameters were tuned in
Network (DSN) tracking of spacecraft will come flight, typical autonomous cruise heliocentric orbit
from the placement of a complete navigation determinations differed from radiometric solutions
capability on board the spacecraft. (Other (developed to provide a reference against which to
autonomy technology experiments are discussed test AutoNav) by < 1000 km and < 0.4 m/s. With
below.) In addition, autonomous navigation simple ground-based removal of some images
allows a smaller navigation team during flight. (based on an algorithm that would be straight-
forward to implement in the flight software),
One portion of the core of the autonomous accuracies improved to < 400 km and < 0.2 m/s.
4
For encounters, navigation is target-relative, 256 × 256 18-µrad/pixel CMOS active pixel
and 1 delivery accuracy is ~ 3 km. AutoNav sensor, which includes the timing and control
also performs target tracking at encounters to electronics on the chip with the detector. The
provide accurate pointing information to ACS, imaging spectrometers operate in push-broom
and it initiates the encounter sequences based on mode. The infrared spectrometer covers the range
its estimate of the time to closest approach. from 1.2 µm to 2.4 µm with spectral resolution of
12 nm and 54-µrad pixels.
During periods of IPS thrusting, AutoNav
controls the IPS. It selects the throttle level based MICAS serves three functions on DS1.
on models of SCARLET power generation and First, as with all the advanced technologies, tests
spacecraft power consumption; pressurizes, of its performance establish its applicability to
starts, and stops the IPS; and commands ACS to future space science missions. Second, AutoNav
achieve the attitude needed for thrusting. AutoNav uses the visible channels for optical navigation.
also commands updates to the throttle level and Third, as a bonus, it will collect science data during
spacecraft attitude every 12 hours. the primary mission at the asteroid and at other
encounters if an extended mission is conducted.
During periods of ballistic coast, AutoNav is
given time windows, in each of which it can The ultraviolet channel, designed to operate
execute a trajectory correction maneuver (TCM), between 80 nm and 185 nm, did not function
which it designs autonomously if it has estab- properly and never returned interpretable data.
lished that a TCM is necessary. In most cases, Several tests were conducted to diagnose the
the TCMs are conducted with the IPS. To save problem, and indications are that the malfunction
time during the final 2 days before an encounter is in the signal chain after the detection of the
(and for the purposes of dedicated AutoNav photons.
testing), the hydrazine RCS is used. With either
propulsion system, if thrust is required in an MICAS images and IR spectra revealed
attitude that is prohibited by ACS, the TCM is scattered light. Stray light analysis and dedicated
autonomously decomposed into two allowed tests established the multiple paths responsible.
maneuvers. The scattered light is the result of spacecraft
surfaces directing off-axis light to reflective
Integrated camera and imaging spectrometer components inside MICAS, particularly the
If NASA is to conduct missions with smaller multilayer insulation surrounding the IR detector.
spacecraft, it is essential to have correspondingly The problem is easily avoided for future missions
smaller science instruments. One of the advanced with different mounting of the instrument and
technologies DS1 tested is the Miniature alteration of the internal baffling. Modifications to
Integrated Camera Spectrometer (MICAS), AutoNav significantly increased its immunity to
conceived and developed by a team from the the light, and the flux is sufficiently low that it is
United States Geological Survey, the University not expected to interfere with encounter science.
of Arizona, Boston University, Rockwell, SSG,
and JPL. In one 12-kg package, this derivative of Integrated ion and electron spectrometer
the original concept for a Pluto Integrated Camera Just as MICAS integrates several different
Spectrometer7 includes two panchromatic visible measurement capabilities into one low-mass
imaging channels, an ultraviolet imaging package, the Plasma Experiment for Planetary
spectrometer, and an infrared imaging spec- Exploration (PEPE)8 combines multiple
trometer plus all the thermal and electronic instruments into one compact package. At 5.6 kg
control. All sensors share a single 10-cm- and 9.6 W, PEPE is less than 25% of the mass
diameter telescope. With a structure and mirror of and consumes less than 50% of the power of a
highly stable SiC, no moving parts are required; comparably performing (but more expensive)
the detectors are electronically shuttered. Space- instrument on Cassini. Designed and built by
craft pointing directs individual detectors to the Southwest Research Institute and Los Alamos
desired targets. National Laboratory, PEPE determines the three-
dimensional plasma distribution over its 2.8 sr
The instrument includes two visible detec- field of view.
tors, both operating between 0.5 µm and 1 µm:
a 1024 × 1024 CCD with 13-µrad pixels and a PEPE measures the energy spectrum of
5
electrons and ions simultaneously from 8 eV to 33 as well as in producing 2-way Doppler and
keV per unit charge with at least 5% resolution. ranging data. The DSN also applied the Ka-band
Rather than using moving parts, it electrostatically signals to the validation and improvement of
sweeps its field of view. PEPE measures ion system designs in preparation for upgrading to
mass with a resolution of 5% in the range of 1 to operational use of Ka-band. As the Earth-
500 amu per unit charge. spacecraft range increased, certain tests were
repeated to assure that the transition through
PEPE plays three roles on DS1. It has threshold in a selected Ka-band data rate would be
validated the design for a suite of plasma physics observed. All communication and radiometric
instruments in one small package; it has assisted tests proved to be in good agreement with models
in determining the effects of the IPS on the local or with X-band results for the tests that were
plasma environment, including interactions with enhanced by simultaneous X-band operation.
the solar wind and photoelectrons4; and it makes
scientifically interesting measurements during the Beacon monitor operations
cruise and the encounters. The SDST generates the tones needed for
beacon monitor operations11, conceived to reduce
PEPE made measurements of the solar wind the large demand that would be expected on the
with the IPS on and off, and a very important DSN if many missions were in flight
result is that the data suggest that SEP can be used simultaneously, as envisioned by NASA. In
on future science missions without interfering beacon monitor operations, an on-board data
with the scientific payload. PEPE data showed summarization system determines the overall
Xe+ returning to the spacecraft from the 1 ampere spacecraft health. The system then transmits one
exhaust plume of the IPS and allowed limits to be of four tones to indicate to the operations team the
placed on electrical charging of the spacecraft. In urgency of the spacecraft's need for DSN
January 1999, a favorable alignment of the DS1 coverage. Without data modulation, these tones
and Cassini spacecraft allowed 36 hours of are detected easily with small, low-cost systems,
collaborative solar wind measurements, with the reserving the large, more expensive DSN stations
two spacecraft separated by nearly 0.5 AU. for command radiation and data reception when a
beacon indicates that such services are needed.
Telecommunications technologies The four tones correspond to i) the spacecraft not
DS1 validated a small deep-space needing any assistance because all is well; ii)
transponder (SDST), built by Motorola, and a Ka- informing the ground that the spacecraft has
band solid state power amplifier developed by encountered an unusual but not threatening event,
Lockheed Martin.9,10 Combining the receiver, so a DSN track should be scheduled when
command detector, telemetry modulator, exciters, convenient; iii) alerting the ground that
beacon tone generator (see below), and control intervention is needed to prevent the loss of
functions into one 3-kg package, the SDST allows important data or to assist in resolving a threat to
X-band uplink and X-band and Ka-band the mission, so DSN coverage should occur soon;
downlink. To achieve the SDST's functionality and iv) requiring immediate assistance because the
without a new technology development would spacecraft has encountered a mission-threatening
require over twice the mass and 4 or 5 individual emergency it was unable to solve. In each case,
subassemblies. The SDST, along with when tracking is initiated, the data summarization
SCARLET and AutoNav, had to function system provides a synopsis of the pertinent
correctly from the beginning of the mission. spacecraft data.
Based on extensive routine use and dedicated
experiments, its performance was in excellent This artificial intelligence technology uses
agreement with preflight tests. adaptive alarm limits, which allow tighter
monitoring than traditional limits. Furthermore,
The SDST's Ka-band signal is amplified by the spacecraft parameters that are monitored and
the 0.7-kg power amplifier to 2.3 W with an their limits depend upon the spacecraft activity.
overall efficiency of 13%. In addition to The system adaptively filters data so instead of
characterizing the operation of the hardware using fixed limits, it can compute variable limits
device, DS1 provided Ka-band signals for DSN on the fly; it can apply this not only to single data
use in verifying systems for acquiring, parameters but also to functions of multiple data
demodulating, decoding, and processing telemetry parameters. These alarm limit functions are
6
"trained" using a neural network on the ground new technology on a dedicated test mission. The
with actual DS1 engineering data to create bug proved to be easily correctable for future uses
functions that can perform more precise anomaly of the technology. Analysis showed that it was
detection and detect important trends sooner than safe to continue experiments on DS1 without
with conventional limits. Although this ground fixing it, so a second experiment was devised,
software is quite complex, only the resulting and it captured the remaining remote agent test
functions are uploaded to the spacecraft. objectives.
Experiments conducted during DS1 Microelectronics and structures
addressed both the data summarization and the Electronics mass, volume, and power
tone generation and detection (in both X-band and consumption are important drivers for overall
Ka-band), which agreed well with preflight spacecraft design. DS1 included tests of two
models. Beacon monitor operations may be relied microelectronics technologies and a mechanical/
upon during an extended mission if it occurs. electronic experiment intended to contribute to the
achievement of NASA's vision of spacecraft in
Autonomous remote agent the future. To reduce the power consumption of
For the third autonomy technology DS1 electronics, one experiment used devices with
tested, an artificial intelligence system was placed very low voltage and low capacitance.13 This low-
on board to plan and execute spacecraft power electronics experiment contained four
activities.12 The team that developed this system ring oscillators and some discrete transistors to
was drawn principally from JPL and NASA's test 0.9-volt logic and 0.25-µm gate lengths
Ames Research Center. Rather than standard (achieved with 248-nm lithography) based on
remote control, this technology uses an agent of silicon-on-insulator technology. Provided by the
the ground team on board the spacecraft. This Massachusetts Institute of Technology Lincoln
remote agent was tested in a restricted case on Laboratory, the functioning of the devices in flight
DS1, in preparation for more ambitious was in good agreement with prelaunch tests. DS1
experiments on subsequent flights. The remote also tested two power actuation and switching
agent includes an on-board mission manager that modules, the result of a joint development among
carries the mission plan, expressed as high-level Lockheed Martin, Boeing, and JPL.14 Each
goals. A planning and scheduling engine uses the device contained four power switches, controlled
goals, comprehensive knowledge of the spacecraft by a mixed-signal ASIC, providing voltage and
state, and constraints on spacecraft operations to current sensing and current limiting. High-
generate a set of time-based or event-based density packaging technology quadruples the
activities, known as tokens, that are delivered to the packing density over the current state of the art.
executive. The executive expands the tokens Designed to be capable of switching up to 40 V
to a sequence of commands that are issued directly and 3 A, the experiment switched an internal test
to the appropriate destinations on the spacecraft. load on DS1. Regular tests showed that the
The executive monitors the response to these performances of both PASMs were consistent
commands and reissues or modifies them if the with prelaunch tests.
response is not what was anticipated. A mode
identification and reconfiguration engine aids in A multifunctional structure15 was provided to
assessing the spacecraft state and in recovering DS1 by the United States Air Force Phillips
from faults without requiring help from the Laboratory and Lockheed Martin Astronautics as
ground except in extraordinary cases. a test panel that was attached to the spacecraft bus.
This new packaging technology integrates
In the experiments on DS1, the remote agent electronic housings and thermal control into load-
operated selected subsystems based on plans bearing elements, thus offering great reductions in
formulated on board. Injection of four the mass of spacecraft cabling and traditional
(simulated) faults tested remote agent's ability to chassis. The DS1 experiment returned data on the
resolve or work around different classes of performance of the electronic connection systems
problems, and in each case it devised the correct for embedded test devices and on the thermal
response. A bug in the executive interrupted the gradients in the panel. The connectors displayed
first experiment, and the successful diagnosis of no evidence of degradation, and the thermal
the problem was one important benefit of the gradients were consistent with preflight
testing; it also illustrated the value of trying out a predictions.
7
MISSION capability exceeded what was needed for DS1,
with relatively low mass and low injection energy
Two objectives provided the impetus for a (in part attributable to the high performance of the
short mission. The principal requirement of DS1 IPS). Including 81.5 kg of Xe and 31.1 kg of
was to return results promptly to the future users hydrazine, DS1 was 486.3 kg at launch, and the
of the technologies. Except for tests of lifetime, Delta provided a C3 = 2.99 km2/s2; the launch
most technologies could be evaluated on short vehicle could have delivered approximately 600
(but intense) missions as well as long ones. In kg to DS1's escape trajectory. The excess launch
addition, in general shorter missions are less vehicle performance allowed the manifesting of
expensive that longer ones. As a result, it was another spacecraft on this launch. SEDSAT-1,
decided that the primary mission would be about built by the Students for the Exploration and
one year. This allowed sufficient time to exercise Development of Space at the University of
all the technologies under a wide range of Alabama in Huntsville, in collaboration with
conditions while keeping costs low and not NASA's Marshall Space Flight Center and
forcing eager potential users to wait unreasonably Johnson Space Center, was mounted on the
long before being confident about the new second stage, which accomplished insertion into
systems. It also allowed sufficient time to Earth orbit. After the second stage's second burn,
accomplish the objective of thrusting with the IPS to raise the orbit of the third stage and DS1, the
long enough to place DS1 on a ballistic trajectory stage separated and carried SEDSAT-1 to its
to an asteroid (the third criterion for success). intended orbit, where it was separated. The third
stage completed DS1's injection to heliocentric
DS1 was planned for launch in July 1998, orbit.
based on the earliest expected spacecraft readiness
in a schedule that was extremely aggressive Following launch, several days were spent
(particularly given the large number of unproven conducting an initial evaluation of the spacecraft,
technologies incorporated into the mission). The verifying its health and preparing it for early
mission design, including solar system encounter mission operations. Dedicated technology
targets, was based on that plan.16 experiments began within one week of launch. Of
course, some technologies were used as part of
In the spring of 1998, it became clear that regular spacecraft operations, in particular the
launching DS1 in its planned launch period solar array, transponder, and AutoNav, but those
presented an unacceptable risk to mission success, and all other technologies also were subjected to
so the launch period was shifted to October - in-depth characterization tests.
November 1998. DS1 was given the slot vacated
by the Far Ultraviolet Spectroscopic Explorer Radiometric determination of the actual
when its launch was moved to 1999. Still, an trajectory was combined with results of the first
unusually dense schedule of launches and other SCARLET and IPS tests to generate and optimize
activities at the Eastern Test Range made sched- an updated low-thrust trajectory that was
uling DS1's launch difficult. Once the launch transmitted to the spacecraft. After verification of
period was selected, a new mission was designed its functional capability, AutoNav was tuned in
with the requirements that it necessitate changes in flight, particularly to account for discrepancies
neither the spacecraft nor ground systems and still between the predicted and the actual MICAS
be compatible with the secondary payload (see images. As the mission progressed, more reliance
below). The original mission plan was sufficiently was placed on AutoNav, with conventional radio
robust that its architecture did not need to be navigation used to validate its performance.
changed, but the encounter targets and
thrusting and coasting times did change. Initial IPS thrusting was conducted with the
thrust vector along the Earth-spacecraft line to
DS1's launch occurred at 12:08:00.502 UTC maximize communications rates and the Doppler
on October 24, 1998. It was launched on the first signature, in order to quantify the actual thrust at
Delta II 7326-9.5 (from The Boeing Company), selected throttle levels. After 10 days of
the smallest vehicle in the Delta stable, and was thrusting, the spacecraft was turned to thrust
the first launch of NASA's Med-Lite program. along the optimal vector (subject to a variety of
This launch vehicle was selected largely on the pointing constraints) for reaching the encounter
basis of prompt availability and low cost, but its targets for the primary and extended mission.
8
To meet the demanding schedule prior to have a quiet cruise. Because of the experimental
launch, some software development was nature of the spacecraft and the technologies, early
completed after launch. The launch load did not sequence development was confined to
include all functions needed to conduct tests with implementing and validating command activity
the low-power electronics, power actuation and blocks that could be modified readily and executed
switching modules, multifunctional structure, on board by real-time commanding to achieve a
beacon monitor operations, and remote agent. desired technology experiment. In the first three
These technologies were selected for exclusion months of flight, about 1800 real-time commands
from the earlier software because they were not were executed by the spacecraft.
needed for the basic operation of the mission. In
February 1999, a completely new software load The judicious use of multimission tools and
of 4.1 megabytes was installed. This new services and standards such as CCSDS
software enabled the testing of four of the (Consultative Committee for Space Data
previously excluded technologies (remote agent Standards) kept the cost of the ground system and
was not in this load), upgraded AutoNav mission operations to a minimum. The small
(principally to accommodate scattered light in the operations team averaged about 50 full-time
MICAS images), corrected bugs identified after equivalents, including system and subsystem
launch, and improved spacecraft operability. analysts, flight controllers, technology support
teams, testbed engineers, and project management
To accommodate the remote agent and staff.
experiments in May, the flight software was
patched; in addition, the remote agent software During the routine IPS thrust periods, one
was uploaded. In June, following the remote DSN pass each week allowed high-rate
agent experiment, the entire flight software was commanding and return of spacecraft engineering
replaced again. This last load contained new and technology validation data through the high
operational enhancements and upgrades to a gain antenna. This weekly track was immediately
number of systems, but primarily it included preceded by AutoNav's collection of optical
further AutoNav upgrades for enhanced image navigation images, and both activities were
processing (such as image differencing to gain conducted with the IPS off. The IPS thrusted for
greater suppression of scattered light and more the remaining 90% of the week. One or two
powerful corrections for MICAS' large geometric shorter passes were scheduled between the longer
distortions) and the functions needed to execute ones. Conducted only with one of the low gain
encounters. All three changes to the flight antennas, to allow communication in the preferred
software, which included substantial development thrust attitude, the shorter passes were used to
and testing, large uplink volumes, and rebooting verify that the IPS was thrusting. On occasion
of the nonredundant main spacecraft computer, this coverage also was used to conduct IPS or
were completed without incident. SCARLET experiments.
The mission operations system made The strategy for selecting IPS thrust and
extensive use of standard tools and mission coast times was based on compromises between
services JPL provides to support a wide range of optimizing the trajectory and conducting the
missions. DS1 employed JPL's multimission technology experiments and other mission
ground data system to provide the uplink and activities incompatible with the attitudes or other
downlink data transport capabilities as well as spacecraft states required for thrusting.16 As
much of the telemetry processing and display illustrated in Figure 1, the deterministic thrusting
system. Project-developed applications for the primary mission was accomplished in two
augmented the system to be consistent with the major periods. The brief hiatus in the first major
autonomous capabilities of the spacecraft. thrust arc was inserted to allow several days for
activation and initial testing of PEPE in the
DS1 mission operations were significantly absence of the IPS plasma, and SDST and Ka-
different from that of typical deep space missions band experiments incompatible with the IPS thrust
at JPL. This was primarily attributable to the attitude. When the second thrust segment ended
technology-validation focus of the DS1 mission. on April 27, 1999 (under direction of AutoNav),
Unlike typical deep space missions, with its very the spacecraft was on a ballistic trajectory that
active technology testing campaign, DS1 did not would encounter asteroid 1992 KD, thus
9
Figure 1. DS1 trajectory for the primary mission (through September 18, 1999) and proposed extended
mission. The dotted portion is for ballistic coast; the solid portion indicates the IPS thrust is on.
accomplishing the third mission success criterion. control the targeting of the final encounter. The
The thrust plan was developed to maintain the maneuvers prior to 2 days before closest approach
option for an extended mission (see below). will be executed with the IPS, and in the final 2
days the RCS will be used to save time.
On July 29, 1999, the spacecraft will
encounter (9969) 1992 KD at 15.5 km/s. The Because 1992 KD is so faint, it will not be
size and shape are poorly known, but this asteroid detected by AutoNav (using MICAS images) until
is believed to be elongate with a mean radius of about 1 day before closest approach; until the
roughly 1 km. During the final 20 days of the asteroid is detected on board, AutoNav will
spacecraft's approach to the body, AutoNav will continue to use 1992 KD's a priori ephemeris. A
require optical navigation images and trajectory flyby 15 km from the center of the body is
correction maneuvers at increasing frequencies to planned. With an expected navigational delivery
10
accuracy of about 3 km (1), this assures a safe EXTENDED MISSION
but very exciting encounter. The last opportunity
for a trajectory correction maneuver will be 3 If the spacecraft remains healthy and the
hours before closest approach. resources for an extended mission are available,
the DS1 project could conduct a scientifically
During the final approach, AutoNav's exciting mission. With the technology testing
MICAS images will be interspersed with MICAS complete, the extended mission would be devoted
images and spectra collected for science purposes. to comet science. With AutoNav controlling the
The late navigation images will contain informa- IPS, the spacecraft would travel to Comet
tion AutoNav needs to provide rapid updates to its 107P/Wilson-Harrington and Comet
estimate of the position of 1992 KD, critical for 19P/Borrelly.
keeping the asteroid in MICAS' field of view.
The 4 command sequences, sequentially As illustrated in Figure 1, most of the
governing activities during the final 5 minutes extended mission would be devoted to IPS
before closest approach, will be activated by thrusting. By the end of the extended mission,
AutoNav based on its estimates of the time to the spacecraft would have expended essentially all
closest approach. Because MICAS is body-fixed, of its Xe, providing a total of about 4.5 km/s.
the termination of imaging is dictated by when the
angular acceleration of the line of sight to the Because of the reduced mission operations
asteroid exceeds ACS' capability to keep 1992 staff and the increasing geocentric range during
KD in the MICAS field of view. Measurements the extended mission, beacon monitor operations
by PEPE and diagnostics sensors for the IPS will likely would be used to augment the team's ability
continue through closest approach. to monitor the spacecraft's health. Demand-
access operations have not been implemented by
The asteroid encounter will allow an the DSN however, so that aspect of beacon
opportunity to gather data on the size, shape, monitor operations cannot be implemented.
geomorphology, albedo, and the mineralogy and
compositional heterogeneity of the surface In January 2001, DS1 would reach Comet
material. It may be possible to measure, or at Wilson-Harrington. This comet was lost after its
least constrain, the asteroid's magnetization, discovery in 1949. In 1992, asteroid (4015) 1979
interaction with the solar wind (including VA was recognized to be the same body. It is
sputtering), and outgassing. possible therefore that this comet was seen just as
its activity was terminating. It is considered to be
The spacecraft will point its high gain a dormant comet or a comet/asteroid transition
antenna at Earth about 1 hour after closest object, with an estimated radius of 2 km.
approach to begin returning technology validation
and science data. Although the data return will With DS1's relative speed of 15.8 km/s, the
require several days, the IPS may resume encounter would be similar to the 1992 KD
thrusting as soon as several hours after closest encounter, but it would occur when the comet is
approach. It turns out that with the antenna Earth- near solar conjunction. Although the operations
pointed, the IPS is within 30° of the optimal team would have reduced control authority at that
attitude for thrusting for the extended mission. time, AutoNav would control the trajectory and
For purposes of the extended mission, it is better timing of sequence activations. Of course, there
to thrust in that attitude than to coast. would be sufficient time to incorporate the results
of the final testing of AutoNav at 1992 KD.
The end of the primary mission is on
September 18, 1999. No new technology In September 2001, DS1 would encounter
experiments are planned after the asteroid Comet Borrelly at 17.0 km/s, within days of the
encounter. Following the completion of the return comet's perihelion; this is one of the brightest and
of data, some minor engineering activities will be most active short-period comets. The nucleus is
conducted to prepare for the resumption of long- believed to be a prolate spheroid of about 4 km ×
term thrusting, and then the regular cycle of IPS 2 km with an active surface area of 7% - 10%.
thrusting, interrupted only for weekly acquisition Science data at the comet that could be collected
of optical navigation images and DSN include the structure and composition of the coma
communications, will resume. and tail (including gas, plasma, and dust), the
11
nature of jets and their connection to surface capabilities of the technologies to create new de-
features, the interaction with the solar wind, and signs. Any informed user, seeking to benefit from
the same kind of characterization of the nucleus as the capabilities of these advanced technolo
at the asteroid. gies, now will encounter lower risk and cost by
building upon the successful results of the DS1
The extended mission plan, although devoted project.
to science, illustrates the benefits of the advanced
technologies. If DS1 had used conventional ACKNOWLEDGMENTS
technologies, including a bipropellant propulsion
system (and excluding the fraction of the solar The members of the DS1 mission operations
arrays needed for operation of the IPS), a and technology teams are gratefully acknowledged
transponder similar to that on the Mars Climate for their impressive work, upon which this over-
Orbiter and Mars Polar Lander (launched a few view is based.
months after DS1), and science instruments with
similar capability but without all the innovations Some of the research described in this paper
being tested on DS1, the spacecraft would be was carried out by the Jet Propulsion Laboratory,
significantly more massive. Reaching 1992 KD, California Institute of Technology, under a
Wilson-Harrington, and Borrelly would have contract with the National Aeronautics and Space
required an injected mass of approximately 1300 Administration.
kg compared to DS1's 486.3 kg. And rather than
being able to share a launch on the least expensive REFERENCES
Delta II, the requirements of this hypothetical
mission would have exceeded the capability even
of a dedicated Delta II 7925, the most expensive 1. Rayman, Marc D. and David H. Lehman,
member of that family. The smallest operational "Deep Space One: NASA's First Deep-Space
US launch vehicle that would have met the Technology Validation Mission," Acta
requirements is the Atlas IIA, for which a shared Astronautica 41, p. 289 (1997).
launch would suffice.
2. Rayman, Marc D. and David H. Lehman,
CONCLUSION "NASA's First New Millennium Deep-Space
Technology Validation Flight," Second IAA
The successful flight of DS1 provided an International Conference on Low-Cost Planetary
extensive body of data characterizing the 12 Missions, Laurel, MD, April 16-19, 1996, IAA-
technologies it tested in space. By operating these L-0502.
advanced technologies under actual spaceflight
conditions, the cost and risk to subsequent users 3. Polk, J. E., R. Y. Kakuda, J. R. Anderson, J.
should be greatly reduced, thus allowing rapid R. Brophy, V. K. Rawlin, M. J. Patterson, J.
integration of the important capabilities they offer Sovey, and J. Hamley, "Validation of the NSTAR
into future space and Earth science missions. Ion Propulsion System on the Deep Space One
Another significant benefit of the testing of tech- Mission: Overview and Initial Results," 35th
nologies on DS1 was the experience gained by AIAA/ASME/SAE/ASEE Joint Propulsion
engineering teams. In many cases, the technolo- Conference, Los Angeles, CA, June 20-24, 1999,
gists had not worked on flight projects, and their AIAA-29-2274.
experiences in both development and operations
should prove helpful in their work on future ver- 4. Wang, J., D. Brinza, R. Goldstein, J. Polk,
sions of their technologies. The incorporation of M. Henry, D. T. Young, J. J. Hanley, J.
the technologies into an operational mission Nordholt, D. Lawrence, and M. Shappirio, "Deep
yielded valuable insights into implementation Space One Investigations of Ion Propulsion
issues that would not be expected to arise in Plasma Interactions: Overview and Initial
typical technology development or conceptual Results," 35th AIAA/ASME/SAE/ASEE Joint
mission studies. In addition, spacecraft, ground Propulsion Conference, Los Angeles, CA, June
system, mission planning, and mission operations 20-24, 1999, AIAA 99-2971.
teams discovered the implications of integrating
these new technologies into their designs and, of 5. Stella, Paul M., Donald G. Nieraeth, David M.
course, learned how to take advantage of the Murphy, Michael I. Eskenazi, and John Stubstad,
12
"Validation of the SCARLET Advanced Array on 11. Wyatt, Jay, Rob Sherwood, Miles Sue, and
DS1," Proceedings of the 34th Intersociety John Szijjarto, "Flight Validation of On-Demand
Energy Conversion Engineering Conference, Operations: The Deep Space One Beacon Monitor
Vancouver, British Columbia, August 1-5, 1999, Operations Experiment," Proceedings of the 5th
paper 1999-01-2630. International Symposium on Artificial
Intelligence, Robotics and Automation in Space
6. Riedel, J. E., S. Bhaskaran, S. P. Synnott, S. (iSAIRAS-99), Noordwijk, The Netherlands, June
D. Desai, W. E. Bollman, P. J. Dumont, C. A. 1-3, 1999.
Halsell, D. Han, B. M. Kennedy, G. W. Null,
W. M. Owen Jr., R. W. Werner, and B. G. 12. Nayak, P. Pandurang, Douglas E. Bernard,
Williams, "Navigation for the New Millennium: Gregory Dorais, Edward B. Gamble Jr., Bob
Autonomous Navigation for Deep Space 1," 12th Kanefsky, James Kurien, William Millar, Nicola
International Symposium on Space Flight Muscettola, Kanna Rajan, Nicolas Rouquette,
Dynamics, Darmstadt, Germany, June 2-6, 1997, Benjamin D. Smith, William Taylor, and Yu-wen
paper 88. Tung, "Validating the DS1 Remote Agent
Experiment," Proceedings of the 5th International
7. Beauchamp, P. M., R. H. Brown, C. F. Symposium on Artificial Intelligence, Robotics
Bruce, G-S. Chen, M. P. Chrisp, G. A. and Automation in Space (iSAIRAS-99),
Fraschetti, T. N. Krabach, S. W. Petrick, D. H. Noordwijk, The Netherlands, June 1-3, 1999.
Rodgers, J. Rodriguez, S. L. Soll, A. H.
Vaughan, L. A. Soderblom, B. R. Sandel, and R. 13. D'Onofrio, R. P., C. L. Keast, A. M.
V. Yelle, "Pluto Integrated Camera Spectrometer Soares, P. W. Wyatt, "Low Power Flight
(PICS) Instrument," Acta Astronautica 35, Experiment: Flight Qualifying Sub-0.25-µm Fully
Supplement 1995. Depleted SOI CMOS Technology," 1999 Digest
of Papers, Government Microcircuit Applications
8. Bolton, S. J., D. T. Young, J. E. Nordholt, J. Conference, pp. 238-241.
L. Burch, H. O. Funsten, N. Eaker, and D. J.
McComas, "Plasma Experiment for Planetary 14. Salim, Abbas, Gerhard Franz, Jim Jud, David
Exploration (PEPE)", in Proceeding of Space Hogue, and Gregory Carr, "PASM, The
Technology and Applications International Advanced Power Actuation and Switching
Forum, Albuquerque, NM, 1997, AIP Module as the Building Block for Space
Proceedings 387, ISBN 1-56396-679-4. Micropower Systems," 1999 Digest of Papers,
Government Microcircuit Applications
9. Herman, M. I., S. Valas, W. Hatch, C. C. Conference, pp. 242-246.
Chen, S. H. Zingales, R. P. Scaramastra, L. R.
Amaro, and M. D. Rayman, "Deep Space One 15. Barnett, David M. and Suraj P. Rawal,
Telecommunication Development," 12th Annual "Multifunctional Structures Technology
AIAA/USU Conference on Small Satellites, Experiment on Deep Space 1 Mission," IEEE
September 6-10, 1998, Logan, Utah, paper AES Systems Magazine, January 1999, pp13-18.
SSC98-IV-1.