M018218590

IOSR Journal of Computer Engineering (IOSR-JCE)
e-ISSN: 2278-0661, p-ISSN: 2278-8727, Volume 18, Issue 2, Ver. I (Mar-Apr. 2016), PP 85-90
www.iosrjournals.org
DOI: 10.9790/0661-18218590 www.iosrjournals.org 85 | Page
Smart GNC Scheme for Autonomous Planetary Landing
Ibrahim Mustafa Mehedi
Center of Excellence in Intelligent Engineering Systems (CEIES)
King Abdulaziz University, Jeddah, Saudi Arabia AND
Electrical and Computer Engineering Department, King Abdulaziz University
Jeddah 21589, Saudi Arabia
Abstract: The moon or other nearest planets are important destinations for space science and the smart
landing is key technology for exploring the different planets without fail. Due to the long round-trip delay of
communication with the earth a pinpoint autonomous GNC (guidance-navigation-control) system will be
suitable for precise landing. Hazard avoidance is another key issue for a safe landing of future planetary
missions. Employing sensors and computers onboard, the lander detects hazards in the landing region, makes
quick decision considering mission constraints within exigent time limit and generate trajectory to provide
necessary command for guidance module that transfers the lander to a suitable safe landing site. Real time
onboard terrain mapping Navigation and Guidance is one which has received much attention in the frame of
planetary exploration. The paper illustrates the investigations concerning soft landing for planetary exploration
and finally proposed a smart and autonomous GNC scheme based on real time range image measurements
technique using laser range finder.
Keywords: Planetary landing; Autonomous; Hazard detection and avoidance
I. Introduction
A several lunar or planetary missions were accomplished in last few decades; the guidance-navigation-
control (GNC) technology is getting more important than ever. Because of long round-trip delay of
communication with the ground station is extremely required to have the mission capability to
perform precession autonomous GNC to the selected landing site [7]. With the help of a smart landing technique
it will be possible to reach landing sites even in areas containing hazardous terrain features like craters, rocks or
slopes. 2D and 3D solution of vehicle dynamics for lunar landing mission is developed by the author along with
trajectory generation technique [10 - 14]. Now a scenario of smart GNC scheme is proposed in this paper for
autonomous Planetary landing mission.
The arrangement of this paper is as follows. Outline of precession planetary landing including Earth-
planet orbital transfer, planet orbit insertion and controlled descent are provided in the second section. Third
section describes a summary of historical scenarios for successful soft landing. Traditional autonomous
navigation and guidance methods for planetary landing are described in fourth section. Fifth section
illustratesproposed smart landing technology for safe planetary landing and finally sixth section concludes the
present investigation.
Fig.1. Mission flowchart
Planetary Landing Mission
Earth-Planetary
Orbit Transfer
Trajectory
Correction
Final Descent-Landing

Smart GNCScheme for Autonomous Planetary Landing
II. Smart Landing Mission Outline
From Earth to planetary journey of mission consists three main phases such as transfer, insertion and
descent. After spacecraft is launched it enters in to the planetary circular orbit with the help of several mid-
course trajectory corrections. Before starting descent, the spacecraft executes burning sequences anti-parallel
direction to its velocity in order to perform some important orbit maneuvers and reaches at the point of closest
approach that places the vehicle in to lower planetary orbit. Then the spacecraft will start descent phase to land
at the specified landing site on the planetary surface with in the allowable error range.
At the end of trajectory correction, Hohmann transfer brings the spacecraft to its lower most suitable
orbit. Although Hohmann transfer is most economical but when the semi-major axis of the final orbit is more
than about 12 times greater than that of the initial orbit, a bi-elliptical transfer is even more economical. After
orbital injection, descent starts to accomplish pin-point landing and performs several successive steps within
limited time constraint. Figure 2 shows the different steps during landing. First step is the deceleration step,
from 15 km to 1km altitude above the planetary surface. In this step proportional velocity of the landing vehicle
to the planetary surface will fall from 1 km/s or so lower velocity.Employing sensors and computers onboard the
lander performs terrain mapping, detects hazards in the landing region, and makes quick decision considering
mission constraints. It is a very challenging task to develop 3D map for navigation system which will enable to
estimate the present state of lander and selecting the suitable landing site using information provided by range
imaging. Extracting the hazard free safe landing site by executing developed algorithm the lander could generate
trajectory. Guidance module follows that trajectory performing translation towards appropriate landing site to
avoid hazardous terrain features like craters, rocks or slopes. Due to the fast dynamics of this descent step, the
requirements in terms of computation speed are quite stringent. The fourth step is the free terminal descent step
to conclude the pin-point touchdown episode.
Fig.2. Different sequences of planetary smart landing
III. Successful Landing Missions
Since 1959 several missions of spacecraft landing has been accomplished on other planets and bodies
in the solar system such as Moon, Venus, Mars, Eros, Jupiter, Titan, Comet 9P, Itokawa and Enceladus. Some
of them were successful soft landings and rests are both intended and unintended crash-landings explored by
USSR, USA, Japan.ESA, China and India at different period of time. Exploring the surface of other celestial
body, Moon has got especial attention. Lunar Prospector, the last lunar surface exploration operation in 1999 by
USA, has ended 24th
spacecraft landing mission on Moon since 1962 starting with Ranger 4 to produce hard
impact on the Moon intentionally but it hit on lunar far side due to failure of navigation system. In 1959, USSR
sent the first spacecraft Luna 2 to reach the surface of the Moon, and it impacted the lunar surface west of Mare
Serenitatis near the craters Aristides, Archimeded and Autolycus. USSR completed their 14th
spacecraft landing
mission on Moon by Luna 24 in 1976 with robotic sample return task. Following Luna 2, USSR carried several
unsuccessful landing missions till to achieve first successful soft landing on Moon in 1966 by Luna 9
transmitting first photographic data from the surface of another planetary body to Earth. Just four months after
Earth
Hohmann transfer
Descent start
Trajectory correction
(1) Deceleration
(2) 3D Terrain Mapping by Sensors Data
Processing
(3) Trajectory Generation and
guidance
(4) Free descent

the landing of soviet Luna 9 mission, Surveyor 1 landed on an extraterrestrial body, Moon, as first soft landing
American spacecraft. This spacecraft was launched in 1966 directly in to a lunar impact trajectory and at a
height of 3.4 m above the lunar surface the engines were turned off to let the spacecraft freely touchdown on the
surface. Following Surveyor 1, USA completed 6 manned missions of 11 successful soft landings on lunar
surface while human carried lunar rover during last 3 manned missions to the Moon surface. On the other hand
USSR never carried out successful manned mission towards Moon instead 7 successful unmanned soft landing
missions on lunar surface including 3 robotic sample return and 2 robotic lunar rover missions. In 1993 and
2006, Japan and ESA respectively launched lunar orbiter and both were intentionally impacted on lunar surface
at end of missions. In 2013 – Chang’e3 - first soft-landing vehicle since 1976, and lunar rover is deployed by
China.
In the next month of first successful soft landing on lunar surface by USSR, Venera 3, the first
spacecraft to land on another planet, was launched by Soviet Union to explore the surface of Venus but the
mission was unsuccessful, and crash-landed on Venus on March 1, 1966. Among Venera series of probes to
Venus, Venera 7 was the first man-made spacecraft to successfully land on a different planet to transmit data
from there back to Earth. In 1985 USSR concluded Venus mission performing 10th
successful soft landing by
Vega 2 landed on Venusian surface. For Venus, USA conducted Pioneer mission consisted two
separate launch; Pioneer Venus 1 & 2. Pioneer Venus 2 sent four small probes into the Venusian atmosphere in
1978. None of these atmospheric probes had designed for soft landing except the large probe had a parachute
that was designed to cut loose at a certain altitude. All the probes sustained till impact on Venusian surface, but
only one probe survived for a considerable period after impact.
Mars 2 and Mars 3 are the first probes impacted on Mars launched by Soviet Union. Both of these two
probes were carrying lander and failed upon landing. Mars 3 was the only soviet lander which was able to send
signal for 20 seconds after landing till 1971 and in 1974 Mars 6 lost the contact at landing. The first successful
soft landing was performed by American landers Viking 1 & 2 in 1976 on Mars. To prove the slogan, “faster,
better and cheaper”, USA was able to land successfully Mars Pathfinder and Sojourner Rover in 1997 with
three years of development and at one fifteenth the cost of Viking mission. It executed different investigations
on the Martial soil and lasted almost three months. Mars Polar Lander was launched to touch down on the
southern polar layered terrainbut because of an unknown reason communication was lost just prior to
atmospheric entry in 1999. Beagle 2 was first European lander designed to explore the Martian surface and sub-
surface and carried by Mars Express “mother ship” in 2003 but lost contact after taking attempt for soft landing.
Spirit and Opportunity are two rovers of NASA’s Mars Exploration Rover Mission, landed successfully in 2004
on Mars surface. The latest successful landing mission on Mars was a robotic spacecraft Phoenix under the Mars
Scout Program and landed successfully on May 25, 2008 near the edge of landing ellipse, little far (25-28 km)
from predicted landing position. On November 10, 2008, the mission was declared concluded after NASA’s
engineers were unable to contact the lander.
As one of the large Near Earth Asteroids, Eros was visited by American probe NEAR Shoemaker which
was launched in 1996. Primary goal of the mission was to study from a circular orbit with a radius of 200 km,
gradually it was altered to 35 x 35 km retrograde orbit in December 2000 and finally two months later the probe
landed on the asteroid’s surface using its maneuvering jets at the end of its mission. NASA studied the planet
Jupiter and its moons launching an unmanned spacecraft Galileo in October, 1989 by the space shuttle Atlantis.
A little more than six years later it arrived at Jupiter in December 1995. After 14 years in space including 8
years of service in the Jovian system, the probe was intentionally directed into Jupiter’s atmosphere to impact on
a planetary body. The Huygens probe was an atmospheric entry probe carried to Saturn’s moon Titan launched
in October, 1997 as part of the Cassini-Huygens mission conducted by ESA, USA and Italy. 20 days after
separation from Cassini orbiter, Huygens landed on Titan in January, 2005 near the Xanadu region. To study the
composition of the interior of the Comet 9P, NASA launched impactor, Deep Impact in January 2005. In July,
2005, it successfully impacted on Comet 9P and returned images as late as three seconds before impact. To
return a sample of material from Itokawa, a small near-Earth asteroid, JAXA launched an unmanned spacecraft,
Hayabusa in 2003. Upon arrival at Itokawa, Hayabusa studied the asteroid’s shape, spin, topography, color,
composition, density and history then it attempted to land in November 2005 but the landing was unsuccessful
and sampling chamber was sealed. Finally the spacecraft is scheduled to return to Earth by June 210. Moreover,
Hayabusa was carrying a detachable lander, Minerva mini-lander but it failed to reach the surface due to an
error during deployment.
IV. Investigations for Pinpoint Landing
Landing error ellipse is currently characterized for descent and landing for missions is roughly greater
than 30 x 100 km. But the GNC method is used for such achievement with out considering any terrain
recognition or hazard avoidance mechanism. After extensive search it is found that among successfully flown
automated landing systems for other planets, none had a terrain-relative GNC system for pinpoint landing. MER

is the only mission used descent image motion system to estimate horizontal velocities but it had neither terrain
recognition nor hazard avoidance capabilities. Its image motion estimation system was equipped with a descent
imager, a radar altimeter, an inertial measurement unit and an algorithm to provide a solution for horizontal
velocity estimation problem which helped for air bags deployment during touchdown. The prominent space
agencies like NASA, ESA, JAXA etc. are planning to demonstrate an ambitious capability which is pinpoint
landing to another planetary body. According to the ongoing research of NASA’s Jet Propulsion Laboratory, a
pinpoint landing is characterized to meet future mission requirements delivering a spacecraft to bellow 100
meters of a targeted landing site [3, 8]. In order to achieve this, several new technologies are on the line to be
developed over the coming few years. Some researches are already conducted and some are in planned.
Following discussion presents an overview of these developing technologies.
Topographic recognition and fixation point extraction: Currently vision is the most effective sensory
in precise space exploration which can give us information on structure or shape of observed target. At the end
of last century, researchers of JAXA proposed 2 key technologies; firstly a direct extraction scheme of desired
geographic feature from a gray scale image of a target landing site and secondly a fixation point tracking
scheme extracted from image taken by navigation camera which determines relative position between the target
point and the spacecraft (Misu et al 1999) [1]. Image based Extended Kalman Filter (EKF) is implemented in
this research. In topographic recognition scheme it is shown to achieve feasibility and robustness even without
detailed information about the target albedo and illumination environment. In their method a look-up table is
formed in advance by various indexes of different geographic categories then the calculated index of featured
image will be matched to obtain the exact geographic category. The index of conformity is calculated with the
help of observed reflectance and the normal directions. To determine appropriate target site of landing, a fuzzy
logic is used. In fixation point tracking, author employed block-matching algorithm to estimate relative position
between target point and spacecraft. If there is failure of fixation point tracking then the next image is expected
and a new fixation point is re-extracted in order not to loose the target point.In this investigation Euler angles
are used for spacecraft attitude parameterization. But the use of quaternion rotation increases the robustness
because unlike the Euler angles, the quaternion has no singularities.
Landmark based topography determination: Landmark is defined in terms of local topography and
albedo maps in a local coordinate system with unit vectors in three directions. Firstly data is extracted for a set
of image space locations for a number of landmarks in a number of images. Then a multi-dimensional
estimation for the locations of all landmarks and the camera locations and the orientations for all of the images
are performed by iteration process. While the landmarks are aligned closely enough to within a few percent of
the size of the map, then the map projected brightness of that image is fit to a simple function of slop and that
enabled to determine height [2]
Landmark based spacecraft position estimation during landing: Craters are common on the surface of
planets, satellites, asteroids and other solar system bodies. It is found landforms and generally bowl shaped with
simple and unique geometry which can be considered as ideal landmarks for selecting the targeted landing site.
Yang et al has investigatedcrater basedposition estimation of spacecraft localization for pinpoint landing on
Mars [3]. With in pre selected landing ellipse craters are mapped in terms of landmarks on Earth using orbital
imagery and using these nominal data lander determines landing site onboard. During descent geometric
recognition technique is used to match craters extracted from current images to a previously created database.
From these match landmarks the position and orientation of the spacecraft is estimated with respect to the
surface of the planetary body.
V. Proposed Smart GNC Scheme
Safe landing is the key issue for planetary landing mission on an unknown terrain and to achieve this
hazard is the main obstacle. Either of two design approachesis essential for safe landing in which one is hazard
tolerance and second is hazard avoidance. Preceding lander missions to the surface of the moon and other
planets have used no-autonomous and hazard tolerant designs rather used physical means to protect the payload
from a shock of impact on the surface. Employing airbags by Mars Pathfinder is an example of hazard tolerant
design. In Hazard avoidance the spacecraft needs to use onboard hazard detecting instrumentin the landing zone
and maneuvering capabilities of landing selecting an alternate site if necessary. Therefore, hazard avoidance is
an important aspect of future landing missions to make possible landing even in such areas that are not flat and
hazard free as the nominal selected landing sites of the current exploration missions.
Our smart landing technology proposes an autonomous, high-speed, real-time and environment-
adaptive GNC algorithm toward a designated and scientifically attractive landing site on unknown planetary
body by terrain mapping and hazard classification using laser range finder. Our ongoing developing algorithm
will be able to generate 3D terrain maps and classifies different hazardous object like craters, high slopes,
shadows, boulders, mountain, valleys etc to detect navigable and non-navigable regions.There are some
functional requirements for this new technology.

1. Unlike the previous missions [3] the proposed system must be able recognize completely new terrain of
other planet because future exploration missions envisage landing on planetary surfaces that are not known
a prior.
2. The system should not only be able to detect hazards, but also to decide about area which is deemed to be
secure for landing and guide the spacecraft towards that safe area. Moreover, the system is responsible to
select the reachable landing site. This attainability will be achieved considering mission constraints such as
available onboard propellant and thruster characteristics.
3. The system must accomplish the task under extreme time constraints. For data collecting from range finder,
processing, mapping, identifying the navigable and non-navigable region, other sensors’ data infusion,
decision making, final trajectory generation and lastly provide guidance toward the final destination should
be done in a few second.The phase of spacecraft localization starts between heat shield separation and
powered descent phase. There is a roughly 60 second window of opportunity to send the spacecraft to its
perfect landing site at low gate from high gate [3]. Therefore, it is a very challenging requirement. On
August 25, 2008 Tokyo based Renesa Technology Corporation announced a news regarding successful
fabrication of high speed processor that realizes processing performance of up to 8620 million
instructionsper second (MIPS). However it demands for further investigation to cope with the proposed
smart landing technology.
An integrated simulation tool is being developing which contains different modules that interact
according to the block diagram shown in Figure 3.
Generating 3D Point Cloud: Our approach for smart landing uses lander equipped with 2D laser range finders.
At high gate position range information generates a 3D point cloud onboard and ultimately this cloud models the
terrain.
Terrain Mapping: Terrain map can be generated once pose estimation and range information are available.
Accurate pose estimation technique uses IMU and the attitude quaternion. IMU is the major measurement
element particularly during orbit maneuver and soft landing phase. IMU includes gyros and accelerometers and
it is one of the key units for GNC system. GNC sensor data from other sensors, such as, sun sensors, horizon
sensors or radar can be easily included in the filter to improve attitude and position estimation. Pose information
and laser finder data will be projected into 3D Cartesian space implementing standard geometry.
Terrain Modeling: Our Terrain algorithm is organized based on Markov statistical model and it divides the
terrain into two areas: navigable and non-navigable areas. Lander considers the comparatively flat surface as
navigable areas for safe landing. Non-navigable area classifies the terrain into different sub categories. These are
actually called hazardous area like craters, high slopes, shadows, boulders, mountains, valleys etc to be avoided.
Eventually the maps generated by the algorithm will clearly show the areas where lander can retarget for safe
landing and the areas considered unsafe for landing.
Fig.3. Block diagram of smart landing technology
Segmentation: Due to sensors noise it may turn out that some part of the map is not properly classified and
these errors exist in a small part of the terrain map. Implementing Markov Random Fields segmentation
technique those errors can be fixed [5]. Basically by this Markov technique the probability distribution for each
segmented cell is précised conditionally with the probability distribution of its neighbor cells.

Hazard Avoidance: According to the algorithms to be developed that will estimate the specific locations and
elevations of surface hazards building 3D elevation map. For navigable flat terrains, generated points by the
range sensor are expected to be well aligned, with a minimal variance in altitude. On the other hand, the
hazardous areas are rough terrain where 3D points are not properly aligned even with large variance in altitude.
Before choosing appropriate landing site finally, some maneuvering constraints will be considered like available
onboard propellant and thruster attainability. Keeping everything in account, algorithms will select the safe
landing location and will pass to the lander guidance module that produces a trajectory to the retargeted landing
site.
Guidance: Guidance is the function to transfer the lander from any current state to touchdown. In our
research, an adaptive guidance algorithm will be developed based on E-Guidance [6] to redirect the lander by
selecting a retargeting landing site.
VI. Conclusion
Precisely landing on other planetary bodies at rough and unknown terrain but scientifically significant
is a difficult and risky task. A smart landing technology consisting real time autonomous terrain mapping,
efficiently hazards identification, avoiding hazards and trajectory generation towards safe landing site is crucial.
In this paper, laser range finder based a smart autonomous landing scheme for unknown planetary body
exploration has been suggested. Our proposed algorithm to be developed will be able to build 3D terrain maps
classifying safe landing sites and different specified hazardous objects and finally to make safe pinpoint landing
trajectory to command guidance module considering required mission constraints within challenging time limit.
Acknowledgement
This article was funded by the Deanship of Scientific Research (DSR), King Abdulaziz University,
Jeddah. Therefore, the author acknowledges with thanks DSR financial support.
References
[1]. Misu, T., Hashimoto, T., Ninomiya, K., “Optical Guidance for Autonomous Landing of Spacecraft”, IEEE Transactions on
Aerospace and Electronic Systems Vol. 35, no. 2, April 1999.
[2]. Robert W. Gaskell, “Determination of Landmark Topography from Imaging Data”, 25th Annual AAS Guidance and Control
Conference, Vol. 111, pp. 167-180, 2002.
[3]. Yang, C., Adnan, A. “Land mark based position estimation for pinpoint landing on Mars”, Proceedings of the 2005 IEEE
International Conference on Robotics and Automation Vol. 22, Issue 22, pp. 4470-4475, 2005.
[4]. Nasab, N. M., Analoui, M., Delp, E. J. “Robust and efficient image segmentation approaches using Markov random field models.”
In Journal Electronic Imaging, Vol.12, No.1, pp. 50-58, 2003.
[5]. Kindermann, R., Snell, J. L., “Markov Random Fields and Their Applications.” Contemporary Mathematics, American
Mathematical Society, 1980.
[6]. Cherry, G.W., “E Guidance – A General Explicit, Optimizing Guidance Law for Rocket Propelled Spacecraft”, AIAA, paper no.
64-638, 164.
[7]. Steven N. Rader, Marcum L. Reagan, Barbara Janoijo, and James E. J., “Human-in-the-Loop Operations over Time Delay: NASA
Analog Missions Lessons Learned” Technical report, NASA Johnson Space Center, Houston, TX, 2012.
[8]. Yu, M., Cui, H.T., Tian, Y., “A new approach based on crater detection and matching for visual navigation in planetary landing”,
Advanced Space Research, 53 (12), 1810–1821, 2014.
[9]. Wang, D., Huang, X., Guan, Y., “GNC system scheme for lunar soft landing spacecraft”, Advanced Space Research, 42 (2), pp.
379–385, 2008.
[10]. Mehedi, I. M. and Kubota, T., “Advanced Descent Scheme for Lunar Landing”, Automatic Control in Aerospace – Online Journal,
Vol. 4, No.1, 2011.
[11]. Mehedi, I. M. and Kubota, T., “Advanced Guidance Scheme for Lunar Descent and Landing from Orbital Speed Conditions”,
Transaction of Japan Society for Aeronautical and Space Sciences (JSASS), Vol. 54, No.184, pp. 98-105, 2011.
[12]. Mehedi, I. M., Kubota, T., “Trajectory Generation Scheme for Precise and Safe Lunar Landing”, Journal of Space Engineering, Vol.
4, Issue 1, pp.1-13, 2011.
[13]. Mehedi, I. M., Kubota, T. Al-Saggaf, U. M., “3-Dimensional Analytical Solution for Lunar Descent Scheme”, Iranian Journal of
Science & Technology, Transactions of Electrical Engineering, Vol. 39, pp.67-77, 2015.
[14]. Mehedi, I. M., Kubota, T. Al-Saggaf, U. M., “3-Dimensional advanced solution for lunar descent and landing”, Journal of
Vibroengineering, Vol. 17, Issue 1, pp.527-543, 2015.

M018218590

More Related Content

What's hot (20)

Viewers also liked (20)

Similar to M018218590 (20)

More from IOSR Journals (20)

Recently uploaded (20)

M018218590