sign in sign up
<<
Home , Deep Space 1, Discovery One, Frank Drake, HAL 9000

How USRA scientists, engineers, and neutron- star astrophysicists helped pioneer new technologies for autonomous control and navigation. Courtesy: http://www.xray.mpe.mpg.de/~web/psrnav/Courtesy:

On 28 October 2015, the Cassini spacecraft passed about 50 kilometers (30 miles) above the south pole of , one of the moons of that’s presumed to have an ice-covered ocean. During this pass, instruments aboard Cassini sampled water vapor and ice that had been emitted through cracks in the surface of the moon.

To maneuver Cassini for this close approach to a very small moon (500 km in diameter), flight controllers communicated with the spacecraft through NASA’s Deep Space Network (DSN). The DSN sends and receives radio signals to and from spacecraft by means of large antennas in Goldstone, California; Madrid, Spain; and Canberra, Australia. In addition to receiving data from spacecraft in deep space, and transmitting instructions to them, flight controllers send “range codes” as a part of their instructions. When a distant spacecraft receives a range code, it immediately returns it to the DSN. The radio signals to and from the spacecraft travel at the constant speed of light (c = 3 x 105 km/sec), allowing the flight controllers to use the time interval between the sending and receiving of the Top: Artist’s concept of the south pole of Saturn’s moon Enceladus, range codes to compute the distance to the spacecraft. showing the plumes of water vapor and ice being emitted through the The range of the spacecraft was one of three data cracks in its surface. Bottom: the Goldstone antenna (Credits: NASA) Credit: Photobank gallery/Shutterstock.com 3 An ultraviolet spectrometer on on spectrometer An ultraviolet might mission flyby a a region of particularidentify With scrutiny. intense for interest scientists current technology, do with whatever make to have of pre-planned sequence observations has been stored reprogram and cannot on-board more examine of those to any region identified closely the newly With a future RA, of interest. based on be revised plans may hours or information this new With ground- flyby. before minutes a turnaround based control, time of hours and is impractical is time of minutes a turnaround the due to impossible physically speed of light. The RIACS computer scientists and computer The RIACS and JPL at ARC their collaborators their an opportunitywon test to Deep ideas with the flight of NASA’s Space One spacecraft, which was one 1998, October launched on 24 after the launch of Cassini. year Deep Space One had some science an – passing close by objectives – but its and a comet a primary validate to mission was including technologies, dozen new engine, silicone an ion-propulsion sunlight concentrate lenses to solar cells, an autonomous onto and the RA. system, navigation that experiment During a two-day the RA was 1999, May began on 17 primary of Deep Space given control became the first One and thereby artificialto fly software intelligence a spacecraft its onboard and control 1 For example: example: For 2 In the new model of operations, In the new will communicate the scientists science goals directly high-level the space-craft.to The spacecraft will then perform science own its planning and scheduling, into those schedules translate will that they verify sequences, damage the spacecraft,not and them without execute ultimately human intervention. routine The on-board computer software computer The on-board the spacecraft allow that would perform certainto functions thought of was autonomously Agent” (RA) of the as a “Remote human operators. ground-based collaborating scientists Computer the software, of on development and Pell including Drs. Barney of RIACS, Nayak Pandurang future space argued that an RA for reduce needed to missions was spacecraft operations costs, to operations in the ensure robust take and to presence of uncertainty, science of unplanned advantage opportunities. 2012, the round-trip travel time was was time travel the round-trip 2012, 34 hours.about computer 1990s, late to In the mid at NASA’s working scientists the (ARC), Center Ames Research Laboratory (JPL), Propulsion Jet Institute Research and USRA’s Science Computer Advanced for on a collaborate began to (RIACS) the dependence of decrease to way spacecraft operations on ground- and the based flight controllers articulated as their vision DSN. They follows: ager 1 entered ager 1 entered oy V interstellar space on 25 August space on 25 August interstellar At the time, Cassini’s pass was the the time, Cassini’s pass was At in a series of remarkable latest of spacecraft navigation examples the DSN and NASA’s performed by during the decades flight controllers But as of space exploration. humankind reached farther into the and beyond, the radio signals times of long travel distant spacecraft and from to mission posed increasing risks to or failed, success. If a subsystem of a possible failure, signs showed hours before many it could take could be sent to around” a “work the spacecraft controllers from when the on . example, For spacecraft Horizons past New flew the round- July 2015, on 14 communication time for trip travel with the spacecraft about nine was hours. When When Cassini made its close When Cassini made Enceladus, the round- to approach time of the range codes trip travel about three hours. The time was taking data during the for available Enceladus was to close approach of seconds. Thus, the tens a few assisting a critical role DSN played with flight controllers NASA’s place positioning Cassini in the right gather the data at the right time to the plumes of Enceladus. from points needed to accurately accurately to needed points The position. Cassini’s the calculate obtained were data points two other the on board using cameras by of moons other spacecraft locate to of background Saturn against the locations stars known with precisely in the sky. called MAPGEN became the first software to plan the work of robots (the Explorer Rovers) on another planet.

The autonomous navigation system (AutoNav) on Deep Space One did not rely on range determinations via the DSN. AutoNav used onboard cameras to track the path of a few bright against the background field of stars. The known paths of the asteroids through the solar system were combined with the image data from the spacecraft’s cameras to triangulate the spacecraft’s position to within ± 250 km and its velocity to within ± 0.2 m/ sec.6 For a spacecraft in the main , these are relatively Artist’s concept of Deep Space One’s encounter with comet Borrelly (Credit: NASA) large errors compared to what can be achieved with standard ground- operation in deep space with no based navigation techniques, i.e., human intervention. In a second With the successful achievement combining radio tracking from the experiment four days later, the RA of all the desired testing, the DSN with optical data from onboard successfully responded to three experiment completed amid many cameras.7 But the disadvantages simulated faults on the spacecraft. references to HAL 90004 and to of dependency on ground-based The first simulated fault was the Star Trek.5 control and maintenance, “the failure of an electronics unit, and increasing position and velocity the RA successfully diagnosed RIACS scientists, including Nayak, uncertainty with increasing the problem and reactivated the Pell, Dr. Ari Jonsson, and Mr. Kanna distance from Earth, as well as unit. The second was a sensor Rajan, were co-inventors of the RA the large propagation delay and indicating that a device onboard architecture and the three main weakening of the signals at large the spacecraft had failed, and artificial intelligence technologies distances”8 argued for the value of the RA determined that it was the used in RA: a smart executive, a experimentation with autonomous sensor, rather than the device, mode-identification and recovery navigation systems. that had failed. The third simulated fault-diagnosis system, and a fault was a small thruster that had mission planner/scheduler. The The journey toward a more effective stuck in a closed condition. The RA team won the NASA Software- autonomous navigation system for RA responded by switching to an of-the-Year Award for 1999. RIACS spacecraft began in the summer alternate spacecraft control mode scientists continued to develop of 1967 at the Mullard Radio that didn’t use the failed thruster. elements of the work they had Observatory of the After the experiment, the Project done on Deep Space One for use University of Cambridge in the UK, Manager for Deep Space One, Dr. in follow-on NASA missions. For when the graduate student Susan Marc Raymond, reported: example, the planning software Jocelyn Bell discovered . Soon after this discovery, Dr. Frank Drake and his colleagues at Arecibo pointed out the utility of the pulsed signals for space navigation of extra-terrestrial civilizations, though the Arecibo group argued that it was very unlikely that the discovered by Jocelyn Bell was a signal from such a sources. Drake and his colleagues noted, however, that for Earth-based civilizations, “the precise timing of the pulses provides a new time service which may be useful in some circumstances.”9

The basis for an X-Ray navigation system (XNAV)15. SSB stands for Solar System Barycenter13.

of earth-based measurements of from a small (~0.1 m2) on-board pulsar arrival times, the necessity X-ray detector yields a three- for large antenna arrays on the dimensional position accurate spacecraft, long integration times to ~150 km. This accuracy to accumulate enough signal to is independent of spacecraft make precise measurements, and distance from the Earth. Present the dispersion of the radio waves techniques for determining the as they travel through interstellar two spacecraft coordinates other space. than range measure angles and Frank Drake thus degrade with increasing In a 1981 report for the JPL, T. J. spacecraft range. Thus, It wasn’t long before astronomers Chester and S. A. Butman raised navigation using X-ray pulsars and graduate students began the idea of using X-ray, rather will always be superior to present to flesh out this suggestion as it than radio, pulsars for spacecraft techniques in measuring these applied to the use of pulsars for navigation: two coordinates for sufficiently space navigation. G. S. Downs of distant spacecraft. At present, the JPL developed a navigational Approximately one-dozen X-ray the break-even point occurs near 11 method for spacecraft based on pulsars are presently known the of . the use of onboard antennas and which emit strong stable pulses software that would measure with periods of 0.7 to ~ 1000 s. The parenthetical reference the pulse arrival times of three By comparing the arrival times of to Earth-orbiting satellites was radio pulsars.10 Downs noted these pulses at a spacecraft and necessary because X-ray pulsars several potential problems with at the Earth (via an Earth-orbiting can’t be measured from Earth’s his technique, including the satellite), a three-dimensional surface due to the absorption of requirement for a continuation position of the spacecraft can X-rays by the Earth’s . be determined. One day of data Thus, Chester and Butman’s method requires an Earth-orbiting satellite to detect them.

Through the 1980s and beyond, work continued on the possibility of using X-ray pulsars for spacecraft navigation, notably through the efforts of Dr. Kent Wood and others at the U.S. Naval Research Laboratory (NRL). Wood proposed the Unconventional Stellar Aspect (USA) experiment to study the feasibility of X-ray navigation (XNAV) onboard the Advanced Research and Global Observation Satellite (ARGOS). ARGOS was a project of the Space Test Program of the Department of Defense (DoD), which was interested in developing an autonomous spacecraft-navigation system that didn’t depend on the availability of its network of Global Positioning System (GPS) satellites. ARGOS was launched into a low-Earth orbit on 23 February 1999, and USA could explore methodologies for attitude, position, and time determination using a single sensor of simple design and low cost.12

Several groups from around the world also worked on the development of a pulsar-based XNAV system. For example, a group led by Werner Becker of the Max- Planck-Institut für extraterrestrische Physik described an iterative approach:

An initial assumption of position and velocity is given by the planned orbit parameters of the spacecraft (1). The iteration starts with a pulsar observation, during which the arrival times of individual photons are recorded (2). The photon arrival times have to be corrected for the proper motion of the spacecraft by transforming the arrival times (3) to an inertial reference location; e.g., the solar system barycenter (SSB).13 This correction requires knowledge of the (assumed or deduced) spacecraft position and velocity as input parameters. The barycenter corrected photon arrival times allow Top: Iterative determination of position and velocity by a pulsar-based navigation system. (figure 145 ) Bottom: Measuring the phase difference then the construction of a pulse profile or pulse between the expected and measured pulse peak at an inertial reference phase histogram (4) representing the temporal location; e.g., the solar system barycenter (SSB)13. The top profile shows emission characteristics and timing signature of the main peak location expected at the SSB. The bottom profile is the one the pulsar. This pulse profile, which is continuously which has been measured at the spacecraft and transformed to the SSB by improving in significance during an observation, is assuming the spacecraft position and velocity during the observation. If the position and velocity assumption was wrong, a phase shift Δφ is observed. permanently correlated with a pulse profile template (figure 146 ) Credit: NASA/Pat Izzo Credit: NASA Credit:

Array of X-ray concentrators for the NICER/ Zaven Arzoumanian and Keith Gendreau X-ray concentrators (XRCs) for NICER/SEXTANT SEXTANT X-ray Timing Instrument 18

in order to increase the accuracy of at least three different pulsars 56 X-ray Concentrators (XRCs) and of the absolute pulse-phase … . If on-board clock calibration associated X-ray detectors. Each measurement (5), or equivalently, is necessary, the observation of a XRC has 24 nested parabolic foils pulse arrival time (TOA). From the fourth pulsar is required.14 to guide X-ray photons onto small pulsar ephemeris that includes silicon detectors by means of the information of the absolute When Dr. Zaven Arzoumanian joined grazing-incidence reflections. The pulse phase for a given epoch, USRA in the fall of 2001, he began energy of an individual X-ray photon the phase difference Δφ between to work with his colleagues at USRA is determined by measuring the the measured and predicted and NASA on various aspects of amount of ionization it produces in pulse phase can be determined. high-energy . About the target silicon. X-rays in the range … In this scheme, a phase shift a decade later, Arzoumanian and 0.2-12 keV will be collected, and for (6) with respect to the absolute Dr. Keith Gendreau of NASA GSFC 1.5 keV X-rays, the total effective pulse phase corresponds to a won an to develop a collection area is nearly 2000 cm2. range difference Δx = cP(Δφ + space mission that, in addition to n) along the line of sight toward some important science objectives The NICER/SEXTANT system offers the observed pulsar. Here c is the related to neutron stars, could the capability to demonstrate speed of light, P the pulse period, provide a means for navigating in for the first time that XNAV can Δφ the phase shift and n = 0, deep space using pulsars. Their determine spacecraft positions ±1, ±2, … an integer that takes mission is titled the Neutron Star with greater accuracy than other into account the periodicity of Interior Composition Explorer/ existing systems. The ISS orbit as the observed pulses. If the phase Station Explorer for X-ray Timing and determined by XNAV via NICER/ shift is non-zero, the position and Navigation Technology, or NICER/ SEXTANT will be compared with velocity of the spacecraft needs SEXTANT. Gendreau is the Principal the orbit as determined by the to be corrected accordingly and Investigator (PI) and Arzoumanian Earth’s GPS system. The goal of the iteration step is taken is the Deputy PI for the mission determining the position of the (7). If the phase shift is zero, or team, which includes scientists and ISS to within a few kilometers falls below a certain threshold, engineers from USRA, GSFC, and will require the measurement by the position and velocity used MIT, as well as other universities NICER/SEXTANT of pulse times-of- during the barycenter correction and NRL. The observing instrument arrival from five to six pulsars to was correct (8) and corresponds for NICER/SEXTANT will be mounted accuracies of 10 microseconds, to the actual orbit of the on the International Space Station where the time needed for the spacecraft. (ISS) in 2017. measurements is less than about 4 hours for each pulsar.15 The need A three-dimensional position fix The NICER/SEXTANT X-ray Timing for measurements from a few can be derived from observations Instrument is a co-aligned set of different pulsars arises because Credit: Aaron Clamage Credit: NASA Credit:

Artist rendition of the NICER/SEXTANT X-ray Timing Instrument mounted on the ISS, shown with sun shades Arzoumanian demonstrates the NICER/SEXTANT (blue) in place technology using a 1/5 scale model. of the “ambiguity problem.” If one and to do it autonomously.16 knew how many pulses intervene between the satellite and the pulsar, If we want to more thoroughly one could calculate the distance to investigate distant planets and the pulsar from the spacecraft by their satellites, XNAV will be a measuring the time between pulses necessity. Earth-centric approaches and multiplying by the speed of light are not only more expensive, but and the number of pulses. The also provide less accuracy than is problem is that one doesn’t know required. The current uncertainties how many pulses intervene between of Earth-based position a given pulsar and the spacecraft. determinations for spacecraft The accompanying figure indicates around planets and their satellites how the measurement of a few beyond Jupiter is in the 10s to 100s pulsars can solve the ambiguity of kilometers, whereas an orbital Solving the ambiguity problem by observing four problem. insertion around Enceladus, for pulsars (drawn in two dimensions). The arrows example, requires an accuracy of point along the pulsars’ lines of sight. Straight lines represent planes of constant pulse phase; 17 As noted by Arzoumanian: 1-5 km. black dots indicate intersections of planes. (figure 147 ) The best current capabilities for At some point in the future, spacecraft position determination autonomous spacecraft control are Earth-centric, resource and navigation will likely be the intensive, and pushed to their norm, made possible in part by the practical limits for critical pioneering contributions of USRA maneuvers at Jupiter and beyond. computer scientists, engineers, and XNAV offers the possibility to astrophysicists. achieve the required accuracies, 1 Pell, B., Bernard, D. E.; Chien, S. A., Gat, E., Muscettola, N., Nayak, P. P., Wagner, M. D., and Williams, B. C. (1996). A remote agent prototype for spacecraft autonomy, In SPIE's 1996 International Symposium on Optical Science, Engineering, and Instrumentation (pp. 74-90). International Society for Optics and Photonics. 2 Bernard, D. E., Dorais, G. A., Fry, C., Gamble Jr, E. B., Kanefsky, B., Kurien, J., Millar, W., Muscettola, N., Nayak, P. P., Pell, B., Rajan, K., Rouquette, N., Smith, B., and Williams, B. C. (1998). Design of the remote agent experiment for spacecraft autonomy. In Aerospace Conference, 1998 IEEE (Vol. 2, pp. 259-281). IEEE. 3 Ibid. p. 260. 4 HAL 9000 is the name of the “Heuristically programmed ALgorithmic computer” in Arthur C. Clarke’s popular novel 2001: A . HAL controlled the systems of the spacecraft. 5 Raymond, M. (1999). Dr. Marc Raymond’s mission log: Voyage of Deep Space 1. Retrieved from http://nmp.jpl.nasa.gov/ds1/arch/mrlogH. html. 6 Riedel, J. E., Bhaskarem, S., Desai, S., Han, D., Kennedy, B., Null, G. W., Synnott, S. P., Wang, T. C., Werner, R. A., and Zamani, E. B. (2000). Autonomous optical navigation (AutoNav) DSI technology validation report. Deep Space 1 technology validation reports (Rep. A01-26126 06- 12), Jet Propulsion Laboratory, Pasadena, California. 7 Becker, W., Bernhardt, M. G., and Jessner, A. (2013). Autonomous spacecraft navigation with pulsars. arXiv preprint arXiv:1305.4842; p. 2. 8 Ibid. 9 Drake, F. D., Gundermann, E. J., Jauncey, D. L., Comella, J. M., Zeissig, G. A., and Craft Jr, H. D. (1968). The rapidly pulsating radio source in vulpecula. Science, 160(3827), pp. 503-507. 10 Downs, G. S. (1974). Interplanetary navigation using pulsating radio sources. 11 Chester, T. J., and Butman, S. A. (1981). Navigation using X-ray pulsars. Jet Propulsion Laboratory, Pasadena, CA, NASA Tech. Rep. 81N27129; p. 22. 12 Wood, K. S. (1993). Navigation studies utilizing the NRL-801 experiment and the ARGOS satellite. In Optical Engineering and Photonics in Aerospace Sensing (pp. 105-116). International Society for Optics and Photonics. 13 Solar System Barycenter (SSB) – the SSB is the center of mass of the solar system, the point about which all members of the solar system, including the Sun, revolve. Primarily because of the mass of the planet Jupiter, the SSB is outside the photosphere of the Sun. 14 Op. cit. Becker et al., 2013, pp. 7-8. 15 Arzoumanian, Z. (2010). Spacecraft navigation via X-ray timing of pulsars: A technology demonstration proposal for International Space Station (ISS) utilization, Dr. Keith Gendreau, Principal Investigator, a presentation given to the Center for Research and Exploration in Space Science and Technology. USRA Archives. 16 Ibid. 17 Ibid. 18 Ray, P. S., Sheikh, S. I., Graven, P. H., Wolff, M. T., Wood, K. S., and Gendreau, K. C. (2008, January). Deep space navigation using celestial X-ray sources. In Proc. ION 2008 National Technical Meeting (pp. 101-109). Figure 2.

Credit for photo of Frank Drake. http://www.skyandtelescope.com/astronomy-news/the-chance-of-finding-aliens/