US008152699B1

(12) United States Patent (10) Patent No.: US 8,152,699 B1

Ma et al. (45) Date of Patent: Apr. 10, 2012

(54) APPARATUS AND METHOD FOR 6,886,471 B2 5/2005 Rodnunsky

REDUCED-GRAVITY SIMULATION 7,780,573 B1* 8/2010 Carmein ........................... 482/4
2003. O153438 A1 8, 2003 Gordon et al.
2004, OO63550 A1 4/2004 Harris
(75) Inventors: Ou Ma, Las Cruces, NM (US); Jiegao 2005/0101448 A1* 5/2005 He et al. .......................... 482.54
Wang, Wuhu (CN) 2006/0241539 A1 * 10/2006 Agrawal et al. ................. 602/23
2006/0260621 A1* 1 1/2006 Agrawal et al. ............... 128,845
(73) Assignee: Arrowhead Center, Inc., Las Cruces, 2007/O123997 A1 5, 2007 Herr et al.
NM (US) 2008/0054836 A1 3/2008 Rodnunsky et al.
(Continued)
(*) Notice: Subject to any disclaimer, the term of this
patent is extended or adjusted under 35 FOREIGN PATENT DOCUMENTS
U.S.C. 154(b) by 58 days. SU 450090 11, 1974
(21) Appl. No.: 12/487,579 (Continued)
(22) Filed: Jun. 18, 2009 OTHER PUBLICATIONS
O O "NASA microgravity flight history'. http://jsc aircraftops.jsc.nasa.
Related U.S. Application Data gov/Reduced Gravity/C 9B history.html Mar. 25, 2008.
(60) Royal application No. 61/074,107, filed on Jun. (Continued)
51) Int. C Primary Examiner — Allana Lewin
(51) E. 3200 2006.O1 (74) Attorney, Agent, or Firm — Deborah A. Peacock;
A63B 22/02 3:08: Samantha A. Updegraff: Peacock Myers, P.C.
(52) U.S. Cl. ............................................ 482/69; 482/54 (57) ABSTRACT
(58) Field of Classification Search .................... 482/54,
482/66–69, 143, 144; 601/5: 473/131; 434/255 The present invention comprises an apparatus and method for
See application file for complete search history. gravity-balanced apparatuses for training humans for space
exploration and other applications. The embodiment of the
(56) References Cited simulation apparatus is less expensive to build and safe to
operate and adaptable to numerous applications, including
U.S. PATENT DOCUMENTS but not limited to theme parks, museums, training facilities,
2,812,010 A * 1 1/1957 Abdallah ........................ 482.69 educational/research labs, and others, for people to experi
3,330,052 A 7, 1967 Johnson et al. ence walking and other perambulations in lower or Zero grav
3,449,843 A 6, 1969 Richter et al. ity environments. The present invention is statically-balanced
3,516,179 A 6, 1970 Dane and comprises a spring apparatus that is easily adjusted. An
3,550,585 A 12/1970 Wayne et al. embodiment of the present invention provides an apparatus
3,583,322 A 6/1971 Vykukal and method for simulating walking in a Zero-gravity or
3,701,528 A 10/1972 Ryan
4,620,829 A 11, 1986 Herve reduced-gravity environment.
4,779,712 A 10, 1988 Peterczak et al.
5,667,461 A 9, 1997 Hall ................................ 482.69 19 Claims, 13 Drawing Sheets
 US 8,152,699 B1
Page 2

U.S. PATENT DOCUMENTS Homan, David J. et al., “An Integrated EVA/RMS Virtual Reality
2009, 0215588 A1* 8, 2009 Riener et al. ...................... 482.7 Simulation, Including Force Feedback, for Astronaut Training”. Pro
ceedings of the AIAA Flight Simulation Technologies Conference,
FOREIGN PATENT DOCUMENTS San Diego, CA. Technical Papers (A96-3500 1 09-01), Reston, VA.
1996, 216-223.
SU 1759.731 9, 1992 Kazerooni, H. et al., “A New Architecture for Direct Drive Robots',
OTHER PUBLICATIONS Proc. of the IEEE Int. Conference on Robotics and Automation Phila
delphia 1988, 442-445.
Agrawal, Abhishek et al., “Effect of Gravity Balancing on Biped Lowen, G. G. et al., “Balancing of Linkages—An Update'. Mecha
Stability”. Proceedings of the 2004 IEEE International Conference nism and Machine Theory vol. 18, No. 3, Pergamon Press Ltd., Great
Britain 1983, 213-220.
on Robotics & Automation New Orleans 2004, 4228-4233. Ma, Ou et al., “Dynamically Removing Partial Body Mass. Using
Bagci, Cemil, "Shaking Force Balancing of Planar Linkages with Acceleration Feedback for Neural Training”. Proceedings of the
Force Transmission Irregularities Using Balancing Idler Loops'. 2007 IEEE 10th International Conference on Rehabilitation Robot
Mechanism and Machine Theory vol. 14, Pergamon Press Ltd., Great ics Noordwijk, The Netherlands Jun. 12, 2007, 1102-1 107.
Britain 1979, 267-284. Nathan, R. H., “A Constant Force Generation Mechanism'. Trans
Banala, Sai K. et al., “Gravity-Balancing Leg Orthosis and Its Per actions of the ASME vol. 107, ASME Dec. 1985, 508-512.
formance Evaluation', IEEE Transactions on Robotics vol. 22, No. 6 Schwartz, Jana L. et al., “Historical Review of Air-Bearing Space
Dec. 2006, 1228-1239. craft Simulators'. Journal of Guidance, Control, and Dynamics vol.
Cardoso, Luis F. et al., "Conceptual Design of a Passive Arm 26, No. 4 Jul.-Aug. 2003, 513-522.
Smith, M. R. "Optimal Balancing of Planar Multi-Bar Linkages'.
Orthosis'. Proceedings of DETC '02 ASME 2002 Design Engineer Proceedings of the 5th World Congress on the Theory of Machines
ing Technical Conferences and Computer and Informtaion in Engi and Mechanisms New-Castle-Upon-Tyne 1975, 142-149.
neering Conference Montreal, Canada Sep. 29-Oct. 2, 2002, 747 Stevensen, Jr., E. N. “Balancing of Machines'. ASME Journal of
756. Engineering for Industry vol. 95, No. 2 Nov. 1973, 650-656.
Ebert-Uphoff, Imme et al., “Static Balancing of Spatial Parallel Plat Streit, D. A. et al., “Perfect Spring Equilibrators for Rotatable
form Mechanisms-Revisited'. Journal of Mechanical Design vol. Bodies”,Journal of Mechanisms, Transmissions, and Automation in
122, ASME Mar. 2000, 43-51. Design vol. 111 Dec. 1989, 451-458.
Fattah, Abbas et al., “On the Design of a Passive Orthosis to Gravity Ulrich, Nathan et al., “Mechanical Design Methods of Improving
Balance Human Legs'. Transactions of the ASME vol. 127, ASME Manipulator Performance'. Proc. 5th Int. Conf. On Advanced Robot
Jul. 2005, 802-808. ics (ICAR 91) vol. 1, Pisa, Italy Jun. 19, 1991, 515-520.
Feng, Gao, “Complete Shaking Force and Shaking Moment Balanc Ulrich, Nathan et al., “Passive Mechanical Gravity Compensation for
ing of 17 Types of Eight-Bar Linkages Only with Revolute Pairs'. Robot Manipulators'. Proceedings of the 1991 IEEE International
Mech. Mach. Theory vol. 26, No. 2, Pergamon Press 1991, 197-206. Conference on Robotics and Automation Sacramento, California Apr.
Fujii, Hironori A. et al., “Ground-Based Simulation of Space 1991, 1536-1541.
Manipulators Using Test Bed with Suspension System”, Journal of Van Dorsser, Wouter D. et al., “Gravity-Balanced Arm Support with
Guidance, Control, and Dynamics vol. 19, No. 5 Sep.-Oct. 1996, Energy-Free Adjustment”. Journal of Medical Devices vol. 1, ASME
985-991. Jun. 2007, 151-158.
Gosselin, Clement M. et al., “Static Balancing of Spatial Six-Degree Walsh, G. J. et al., “Spatial Spring Equilibrator Theory'. Mech.
of-Freedom Parallel Mechanisms with Revolute Actuators', Journal Mach. Theory vol. 26, No. 2, Pergamon Press, Great Britain 1991,
of Robotic Systems vol. 17. No. 3, John Wiley & Sons, Inc. 2000, 155-170
159-170 Wang, Jiegao et al., “Static balancing of spatial three-degree-of
Griffin, B. N., “Zero-G Simulation Verifies EVA Servicing of Space freedomparallel mechanisms’. Mechanism and Machine Theory vol.
Station Modules'. ALAA Space Station in the Twenty-First Century 34, Elsevier Science Ltd. Jul. 1999, 437-452.
AIM-86/2312, Reno Nevada Sep. 3, 1986, 2-5. Ye, Z. et al., “Complete Balancing of Planar Linkages by an Equiva
Herder, Just L., “Development of a Statically Balanced Arm Support: lence Method”. Mech. Mach. Theory vol. 29, No. 5, Elsevier Science
ARMON”. Proceedings of the 2005 IEEE 9th International Confer Ltd. Great Britain 1994, 701-712.
ence on Rehabilitation Robotics Chicago, IL. Jun. 28, 2005, 281
286. * cited by examiner
U.S. Patent Apr. 10, 2012 Sheet 1 of 13 US 8,152,699 B1

37

Fig. 1
U.S. Patent Apr. 10, 2012 Sheet 2 of 13 US 8,152,699 B1

Secondaraieiogas

Fig. 2
U.S. Patent Apr. 10, 2012 Sheet 4 of 13 US 8,152,699 B1

vir'91-)
U.S. Patent Apr. 10, 2012 Sheet 5 of 13 US 8,152,699 B1
U.S. Patent Apr. 10, 2012 Sheet 6 of 13 US 8,152,699 B1

44

41

45

46

47

Fig. 5
U.S. Patent Apr. 10, 2012 Sheet 7 of 13 US 8,152,699 B1

1.
Earth Gravity, =
11 pwars
= 0.5Gravity, p = 0.82
floon Gravity, p = 5.8
E 1-- - - - - - - - - - - - - -
S.
-S, 90
s
I
E 8

E.

Tire (Sec)

Fig. 6
U.S. Patent Apr. 10, 2012 Sheet 8 of 13 US 8,152,699 B1

1.
Earth Gravity,
-----
^^^ - Mars Gravity, p.
''' floon Gravity,

:
S --- ----- - --
S- S. - - - - - - - - - - -
S.
St I
S. s

- --- -------- - --- - - -- - - - - - -- - - - ---- - - - - -

S - - - - -s - - - - - - - - -
R

1.
Tig S.

Fig. 7
U.S. Patent Apr. 10, 2012 Sheet 10 of 13 US 8,152,699 B1

x:
:
3 S.

1
º

14971
s :
3
-8 x x as x w 2: & x 8. as tasketers ea asks:
is s:
88.88x8 it sesses 88-88-888 g :
&
s is sees $88.88

8
:
s

is . . . . . . . . . . x s a ses. . . es: a 9 se e s : x s 8 x 8. x s is es : 8s . . . . . . . g . . . . . . . e. exe & . . . . . . .

0:1
(N) alojo
 US 8,152,699 B1
1. 2
APPARATUS AND METHOD FOR other degrees of freedom are either constrained or do not
REDUCED-GRAVITY SIMULATION match motion as it actually occurs in space. An air-bearing
supported system performs 2-D or preudo 3-D simulation
BACKGROUND OF THE INVENTION only. The virtual reality simulations provide a visual effect
without much real physical reaction. The neutral buoyancy
1. Field of the Invention (Technical Field) technology which is the most commonly used existing tech
The present invention relates to a gravity-balanced passive nology Suffers from water viscous drag, sealing problems,
apparatus and method to physically simulate perambulation and an onerous burden of multiple safety measures.
Such as Walking, hopping, jogging, running, or other move Existing reduced-gravity simulation technologies either
ments in reduced gravity condition. 10
cannot generate a full range of physical motion in space or
2. Description of Related Art cannot be easily or economically accessed. Therefore, there
The design of balancing apparatuses has been an active
research topic for several decades. The problem of static and exists a need for developing alternate methods for reduced
dynamic balancing of linkages has been studied extensively gravity simulation which are inexpensive and which are eas
in the past. Static balancing of an apparatus occurs when the 15
ily implemented.
links and payload do not exert any torque or force on the joints SUMMARY
of the apparatus at any configuration of the apparatus. This
condition is also referred to as passive gravity compensation.
Many gravity-compensated serial and parallel manipula The present invention comprises a reduced-gravity simu
tors have been designed using counterweights, springs and lator assembly comprising: a platform comprising at least five
Sometimes cams and/or pulleys. A hybrid direct-drive grav degrees of freedom, at least one leg exoskeleton comprising at
ity-compensated manipulator has also been developed. More least three degrees of freedom, at least one spring, and a
over, gravity-balanced leg exoskeletons used for assisting treadmill. A plurality of connectors comprise spring attach
persons afflicted with hemiparesis to walk have been studied. ment points. The five degrees of freedom platform comprises
Two basic approaches exist for static balancing, namely, 25 at least one parallelogram, at least one strut, at least one
using counterweights or using springs. An apparatus has the backplate, and at least one body harness. The three degrees of
property of maintaining its mass center at a globally fixed freedom leg exoskeleton Supports and balances a weight of a
location when using the counterweight approach. Static bal human leg, and the five degrees of freedom platform Supports
ancing is achieved in any direction of the Cartesian space of an entire human body weight.
the apparatus. This property is useful for applications in 30 An upper portion of the leg exoskeleton attaches to a
which the apparatus must be statically balanced in all direc human thigh and a lower portion of the leg exoskeleton
tions, e.g. if the apparatus is to be installed in an arbitrary attaches to a human lower leg. The two leg exoskeletons
direction with respect to the gravity acceleration vector. How comprise at least three degrees of freedom. A backplate is
ever, the drawback of this balancing method is that additional disposed adjacent to a human torso and hinged to a strut,
weights must be added to the system which results in larger 35 where the Strut comprises an upper strut portion and a lower
inertia forces because of the added mass of the system. Strut portion.
Alternatively, when springs are used for static balancing, The reduced-gravity simulator further comprises a plural
the total potential energy, i.e. the gravitational potential ity of rotating couplings comprising hinges. The reduced
energy plus elastic potential energy of the apparatus is main gravity simulator further comprises eleven independent
tained constant and the weight of the entire apparatus is 40 degrees of freedom.
balanced with a much smaller total mass than when using The present invention comprises a method of simulating
counterweights. However, a spring-based balancing appara reduced gravity comprising attaching a reduced-gravity
tus balances only along the direction of the gravity vector, simulator assembly to a human, the assembly comprising a
which is unsuitable for some applications. platform comprising at least five degrees of freedom, attach
In space exploration missions, astronauts are often 45 ing at least one leg exoskeleton to a human, the exoskeleton
required to perform extra-vehicular activities (EVA). Such comprising at least three degrees of freedom, attaching at
activities occur either in a microgravity environment Such as least one spring to the simulator, and disposing a treadmill
on the International Space Station (ISS) or in a reduced adjacent to a human; adjusting spring attachment points
gravity environment such as on the Moon or Mars. To ensure achieving gravity balancing; and reducing gravity to a desired
the Success of a mission, extensive training is required for the 50 level. The present invention comprises a method further com
astronauts before a real mission is launched. Astronauts usu prising using the treadmill to allow continuous forward walk
ally spend more than ten times the real EVA time in a ground 1ng.
based microgravity training facility Such as a neutral buoy The present invention comprising a reduced gravity simu
ancy pool when practicing a planned EVA task. Therefore, the lator assembly that attaches to the human by attaching a
training technology and facility have a significant impact on 55 human torso to a backplate, allowing the torso to freely move
the quality, cost, and time of the required training. in five degrees of freedom, Supporting at least one human leg
Several existing technologies can be used for EVA training by the three-degrees-of-freedom exoskeleton allowing free
in a simulated reduced gravity condition, Such as a neutral movement of the leg, and providing redundant degrees of
buoyancy pool, parabolic-trajectory flight, counter-weight freedom about the Z axis to allow ergonomic comfort while
Suspension; air-bearing/gimbal Support, and virtual reality. 60 the human is walking.
All of these technologies have drawbacks when used for The present invention comprises a method for adjusting
physical simulation of reduced gravity conditions. For springs for gravity balancing comprising sliding the spring
example, the parabolic-trajectory flight technique can simu attachment points and measuring pressure until the measured
late Zero-G for only 20 to 30 seconds and thus is too brief for pressure underneath the human foot is completely gone but
training most of the EVA tasks. The counterweight balanced 65 the foot still touches the treadmill, changing the body and/or
Suspension method can effectively provide only one-degree leg pose, and repeatedly sliding the spring attachment points
of-freedom controlled motion in the vertical direction. The until required balancing is achieved. The pressure is mea
 US 8,152,699 B1
3 4
Sured manually or measured automatically by installing an components. The present invention is inexpensive to build,
actuated spring adjustment device for each spring. easy to operate, and safe to use.
The method of the present invention further comprises the The present invention uses the principle of maintaining
steps of Scaling the spring attachment points based on the constant potential energy to statically balance the gravity
percentage of gravity which is to be reduced and retaining a 5 force at any working configuration of the apparatus. The
percentage of the pressure under a foot of the human because apparatus balances full or partial gravity simply by adjusting
partial gravity is still present. the stiffness values or the attachment locations of the springs.
It balances not only the gravity force applied on the torso of
BRIEF DESCRIPTION OF THE DRAWINGS the human Subject but also the gravity forces applied to the
10 Subject’s legs at any leg configuration. Adding a treadmill, the
The accompanying drawings, which are incorporated into apparatus easily simulates free and continuous walking, run
and form a part of the specification, illustrate one or more ning, gravity
and jumping/hopping activities in a Zero or reduced
condition. The present invention has better dynamics
embodiments of the present invention and, together with the performance because it eliminates actuators and is made of
description, serve to explain the principles of the invention. 15 lightweight materials.
The drawings are only for the purpose of illustrating one or The present invention comprises a reduced-gravity appa
more preferred embodiments of the invention and are not to ratus, which uses spring-based gravity-balancing technology
be construed as limiting the invention. In the drawings: to simulate human walking or jumping in a reduced-gravity
FIG. 1 is a schematic representation of an embodiment of environment. The apparatus is capable of assisting astronauts
the passive apparatus based Zero-G or reduced-G training 20 training for EVA. The reduced-gravity system of the present
facility of the present invention; invention compensates for the full or partial gravitational
FIG. 2 is a schematic illustration of the static-balancing effect of the trainee, providing a similar experience or feeling
training apparatus of the present invention; as in a real reduced-gravity environment.
FIG. 3 is a schematic illustration of an adaptive reduced-G The invention has been studied and developed by means of
simulation facility; 25 simulation and nonhuman experiments. In the simulation, a
FIG. 4 is a simulation model of an embodiment of the physical human is modeled as a multi-body dynamical sys
present invention implemented on MSC ADAMSTM software tem of 54 degrees of freedom. The dynamic responses of a
for simulation study; human walking or jumping with the reduced-gravity appara
FIG. 5 is an illustration of a human skeleton model imple tus are simulated and analyzed. The simulation results are
mented on LifeMODTM for simulation-based study of the 30 compared to those of the same human body on free walking
present invention; and jumping in the reduced gravity environment.
FIG. 6 is a graph illustrating the vertical displacements of A reduced-gravity simulation approach for astronaut train
the center of body mass of a human jumping with the RGS ing is illustrated in a simulator shown in FIG. 1, and is based
under different levels of gravity; on a static balancing technology. Static balancing of an appa
FIG. 7 is a graph illustrating the vertical displacements of 35 ratus refers to the status that no joint forces or torques are
the center of body mass of a human free jumping under required to keep the apparatus in equilibrium for all configu
different levels of gravitational force; rations within its workspace. The statically-balanced appara
FIG. 8 is a graph illustrating ground reaction force on feet tus of the present invention totally or partially compensates
when a human walks with or without an RGS, and for the gravitational force on an astronaut who is attached to
FIG. 9 is a graph illustrating joint torques in the Sagittal 40 the apparatus to train to walk or do other tasks in a space
plane when a human walks with an RGS. environment. The embodiment of the present invention com
prises a reduced-gravity simulator (RGS) comprising static
DETAILED DESCRIPTION OF THE PREFERRED balancing technology. The present invention allows different
EMBODIMENT levels of gravitational force to be compensated for according
45 to training needs. Lunar activity is simulated by removing %
The present invention comprises an apparatus and method of terran gravity and Mars activity is simulated by removing
for gravity-balanced apparatuses for training humans for 62% of terran gravity.
space exploration and other applications. The embodiment of The term “treadmill', as used throughout the specification
the simulation apparatus is inexpensive to build and safe to and claims is intended to include any surface (including the
operate and adaptable to numerous applications, including 50 ground or an apparatus) useful for leg movement, including
but not limited to theme parks, museums, training facilities, but not limited to a walking, running, jogging or jumping
educational/research labs, and others, for people to experi platform or moving or moveable Surface.
ence walking and other perambulations in lower or Zero grav The term "spring, as used throughout the specification and
ity environments. The present invention is statically-balanced claims is intended to include any spring or spring-like or
and comprises a spring apparatus that is easily adjusted. An 55 elastic device that results in a spring or spring-like motion.
embodiment of the present invention provides an apparatus FIG. 1 illustrates an embodiment of an embodiment of the
and method for simulating walking in a Zero-gravity or training system of the present invention comprising multi
reduced-gravity environment. DOF reduced gravity assembly 10, comprising 5-DOF plat
The present invention relates to methods and apparatuses form 12, two 3-DOF leg exoskeletons comprising slidingly
for using a statically-balanced passive apparatus to simulate 60 adjustable upper portion 14 for Supporting the thigh and slid
walking, running, jumping, hopping, or other movement in a ingly adjustable lower portion 15 for Supporting the lower leg,
Zero or reduced gravity environment. The apparatus is useful and treadmill 37. The 5-DOF platform balances the weight of
for astronaut training and for manned planetary exploration an entire human body and the 3-DOF exoskeletons balance
missions. the two human leg weights. The 5-DOF platform 12 com
The present invention comprises a passive multi-DOF 65 prises first parallelogram 30 and second parallelogram 32,
spring-enforced parallel mechanism, without any hydraulic, Strut comprising upper Strut 34 and lower strut 23, backplate
pneumatic, electrical, magnetic, or other active or powered 35 attached to and Supporting a human torso, and body har
 US 8,152,699 B1
5 6
ness 36 for attaching the human torso to the backplate. Tread manually or automatically. If automation is desired, an actu
mill 37 simulates forward motion while a human is walking ated spring adjustment device is installed for each spring.
on the treadmill. The reduced gravity assembly 10 is disposed Spring attachment points 21 are scaled based on the per
on fixed base 18, and comprises balancing springs 20 and centage of gravity which is to be reduced, if simulation of the
rotating couplings comprising hinges 22 allowing rotation gravitational force of Mars, the Moon, or Earth is desired. The
about vertical axes. Spring attachment points 21 are slidably pressure under the feet of the person will no longer be zero
attached to one point of a first connector 19 and rotationally when the adjustment is done because partial gravity is still
attached to a point 24 of a second connector 19 in the 5-DOF present.
platform 12. Spring attachment points 21 are slidably An embodiment of the present invention comprises a sys
attached to one point of a first connector 19 and fixably 10 tem comprising a 5-DOF platform, two 3-DOF leg exoskel
attached to a second point 26 on upper portion 14 as well as etons, and a treadmill. The 5-DOF of the platform are the
lower portion 15 of 3-DOF leg exoskeleton. three translational degrees of freedom and two rotational
Different gravity forces are preferably balanced by adjust DOF (i.e., the pitch and yaw rotations). The three DOF of the
ing the stiffness values or attachment locations of springs 20 leg exoskeleton are the two rotational DOFs for the hip joint
in the system according to training needs such as removing/6 15 and one DOF for the knee joint. The 5-DOF platform is used
of the gravity for simulating a lunar activity of 62% of the to Support and balance the body weight of a person interacting
gravity for an activity on the surface of Mars. with the system and the 3-DOF exoskeletons are used to
In another embodiment, the apparatus and method use a Support and balance the weights of the two legs of the person.
gravity-balanced apparatus and treadmill to simulate a human The treadmill is used to simulate the forward motion while the
walking in a Zero-G or reduced-G environment. The preferred person is walking.
statically balanced apparatus has nine degrees of freedom, The statically-balanced apparatus comprises three degrees
which are capable of balancing partial or full earth gravitation of freedom in each of the leg exoskeletons and five degrees of
on the ground for any working configuration of the apparatus. freedom in the support platform, and therefore eleven degrees
Preferred static-balancing conditions are derived which are of freedom combined, and is capable of balancing partial or
used for application of a gravity-force balancing system. 25 full earth gravitation on the ground for any working configu
Features of embodiments of the invention comprise: ration of the apparatus. As can be appreciated by those skilled
1) A fully passive spring-enforced parallel apparatus, with in the art, additional or fewer degrees of freedom may be
out any hydraulic, pneumatic, electrical, magnetic, or other added or eliminated to or from the various components. Grav
active or powered components; ity balancing conditions for designing such an apparatus are
2) Maintaining constant potential energy to statically bal 30 derived from the total potential energy of the system.
ance the gravity force at any working configuration; An embodiment of the present invention comprises a
3) Balancing full or partial gravity by adjusting the stiff mechanical method of simulating walking in Zero-gravity or
ness values or the attachment locations of springs; reduced-gravity condition using a passive apparatus compris
4) Balancing not only the gravity force applied on the torso ing a statically-balanced apparatus in which the weight of the
of the human subject but also those forces applied on the 35 moving parts does not produce any force or torque at the joints
Subjects legs at any leg configurations; at any configuration of the apparatus. Therefore, the stati
5) A treadmill that simulates free and continuous walking, cally-balanced apparatus comprises a mechanical system to
running, jumping/hopping or other perambulations in a Zero balance the gravity force of the body and legs of an astronaut
or reduced gravity condition; and or any other person who performs walking training on earth to
6) Better dynamics performance because it has less inertia 40 prepare for space activities.
than other existing devices designed for similar purposes. FIG. 2 is a kinematics illustration of an embodiment of the
Simulating reduced gravity is achieved on the embodiment present invention comprising a gravity-balanced passive
of the present invention by attaching a human to the reduced apparatus for human training. The body harness and leg
gravity simulator assembly for whole body simulation via exoskeletons respectively include the torso and the legs of the
harness 36, adjusting attachment points 21 of springs 20 to 45 human while training with the equipment. The system com
achieve gravity balancing, even when the human mass distri prises a total of eleven degrees of freedom.
bution is unknown. Gravitational force is reduced to any The apparatus excluding the human body has a total of
desired level. eleven degrees of freedom (DOFs) represented by joint
A whole-body simulation comprises attaching the human angles 0-0 as shown in FIGS. 1 and 2. The 0 and 0.
torso to backplate 35, thus allowing the torso to freely move 50 angles do not show up in FIG. 2 because they do not affect the
in five degrees of freedom. The only missing DOF is roll potential energy of the system. Each leg exoskeleton is fixed
motion about the X axis, which does not exist in human on each of the human's upper leg (thigh) and lower leg (crus),
walking. The two human legs are supported by the 3-DOF leg so the three DOFs provided by each leg's exoskeleton are
exoskeletons comprising upper portion 14 and lower portion actually associated with the hip and knee joints of the person.
15, allowing free walking of the legs. Redundant degrees of 55 The hip joint rotates about two axes and the knee joint rotates
freedom are provided about the Z axis in order to allow about one axis, as indicated in FIG.1.
ergonomic comfort while the human is walking. Treadmill 37 To analyze the apparatus, an inertial reference frame
allows continuous forward walking in unison with reduced X-Y-Z is fixed to the ground with the Y-axis pointing into the
gravity system 10. paper as depicted in FIG. 2 and the Z-axis pointing vertically
Springs 20 are adjusted for balancing gravity by attaching 60 upward, and the X-axis pointing to the right. The angle 0.
the human to reduced gravity system 10, sliding spring (i=1,2,..., 9) denotes the angular displacement of the ith
attachment points 21 until the pressure underneath both degree of freedom. All the angular displacements are angles
human feet is completely eliminated while the feet still touch measured from the Z-axis to the center lines of the corre
the ground or treadmill. A force plate is used to measure the sponding joints except for the 0-0 angles which are angles
pressure. The human body and/or leg pose is changed and the 65 from the X-axis to the centerline of the link and measured in
spring attachment points are again moved by sliding until the the X-Y plane. These four angles provide extra degrees of
required balancing is achieved. These steps are done either freedom for the human to improve maneuverability but they
 US 8,152,699 B1
7 8
do not affect the analysis because they are independent of the -continued
potential energy of the system. W = (h+ icosé + l2 cosé2 - is + 6)ngg
The torso of the human (the link with the mass m) pref
erably has three translational degrees of freedom along with
the X, Y, and Z directions and two rotational degrees of 5
freedom about the Y and Z axes. Each of the two legs prefer
ably has three rotational degrees of freedom, two with the hip
joint and the third one with the knee joint.
The direction of the gravitational acceleration is preferably
parallel to the negative direction of the Z-axis. Seven springs 10
are preferably used for balancing the weights of the apparatus
and the attached human. The first spring (with stiffness k) is where Volgm 1: Voign2: V body Veg1. Veg2 and Vstrut are,
disposed between the two vertical bars of the first parallelo respectively, the gravitational potential energies of the first
gram apparatus; the second spring (with stiffness k-) is dis parallelogram, the second parallelogram, the body and its
posed between the vertical bars of the second parallelogram 15 harness, the left leg with its exoskeleton, the right leg with its
apparatus. The third spring (with stiffness k-) is disposed exoskeleton and the strut; m and m are the masses of the
between the vertical strut and the body harness of the human. links in the first and second parallelograms; m represents the
The fourth and fifth springs (stiffnesses k and ks) are pref mass of the body and its harness; m and ms are the masses of
erably attached to the upper and lower parts of the leg exosk the upper and lower parts of one leg and its exoskeleton: m is
eleton, respectively. Identical springs are preferably attached
to both leg exoskeletons under the assumption that both legs the mass of the strut disposed to support the weight of the
of the human have the same mass distribution. astronaut. 1 and l are the lengths of the links of the first and
The reduced-gravity apparatus comprises seven springs to second parallelograms; 1 is the distance from the joint of the
compensate the weights of the apparatus and the attached second parallelogram to the joint connecting the strut, the
human. The arrangement of the springs is also shown in FIG. 25 body and the two leg exoskeletons in Z direction: l is the
2 and alternately comprises a plurality of other arrangements length of the upper part of a leg exoskeleton, which comprises
and assemblies of springs due to assembly/manufacturing a parallelogram, r and rare respectively the distances from
requirements. The stiffnesses of the springs are denoted by the pivoting joints to the mass center of the links in the first
parameters k-kz. Among these springs, the first three (with 30 and second parallelograms; r is the distance from the joint
stiffnesses k-k-) are attached to the platform for balancing connecting the body and strut to the mass center of the body;
the weight of the human and the apparatus. The other four r and rs are the distances between the pivoting joints to the
(with stiffnesses ka-k) are installed on the two leg exoskel mass center of the upper and lower links of the leg exoskel
etons to compensate the weight of the two legs and the exosk etons; r is the distance from the joint connecting the strut and
eletons themselves. The two legs of the human are assumed to body harness to the mass center of the strut, and h is the
be identical in both geometrical structure and inertial prop 35 vertical distance from the joint of the first parallelogram to the
erty and thus, the springs on both legs are also identical. X axis; he is the distance between the two links of the first
An apparatus is statically balanced when the total potential parallelogram; his the distance from the joint connecting the
energy contributed by both the gravity and the springs body harness and strut to the attachment point of the spring
remains constant for all working configurations of the appa connecting the body and strut; his the distance from the joint
ratus. This is mathematically equivalent to equation (1), 40 connecting the upper and lower parts of a leg exoskeleton to
where V, V, and Vs represent the gravitational potential the attachment point of the fifth spring; hs is the distance
energy of the apparatus, gravitational potential energy of the between the two links of the leg parallelogram; d, and d are
human, and the spring potential energy, respectively. Under respectively the distances from the joints of the first and
the conditions given by (1), the gravity effect is reduced to 45 second parallelograms to the attachment points of the first and
ZO.
second springs; d is the distance from the joint connecting
the body harness and strut to the attachment point of the third
VIf Yo-Yo-Ys-Constant (1) spring on the strut. d is the distance from the joint of the
Similarly, if only a part of the gravity effect on the human upper part of the leg to the attachment point of the fourth
is required to be removed, the following condition exists, 50 spring; ds is the distance between the two attachment points
where the factor p is defined as the ratio of the removed of the fifth springs.
gravity over the original gravity. Thus, p is also the percentage The portion of each spring shown in FIG. 2 is preferably the
of the gravity to be balanced out. Note that p is a factor of only whole stretch of the spring (the rest of the spring is not shown
the potential energy of the human body because only the in the Figure). Such a spring design is implemented prefer
human requires a reduced gravity while the entire apparatus 55 ably by using cables and pulleys in the spring assembly. The
must have Zero gravity at all times. elastic potential energy of each spring is written as the fol
lowing equation (4), where V, V, and V are the potential
V+pV+Vs-Constant (2) energies of the first, second, and third springs, respectively.
When the X-Y plane possesses Zero gravitational potential V, Vs is the potential energy of the fourth and fifth springs and
energy, the gravitational potential energy of each part of the 60 S is that of the sixth and seventh springs.
apparatus is expressed as follows:
1 2, 12 (4a)
V = ski (di + i - 2dlicosé)
(3)
65 1 2 12 (4b)
V2 = sk:(d; + l; -2dlicosé)
 US 8,152,699 B1
10
-continued The total energy, as given by Eq. (7), is constant for all
1 4 configurations when all the coefficients in the second term of
Vs = k (di + hi-2dsh;costly) (4c) the right-hand side of Eq. (7) become Zero. In other words, the
conditions for partial gravity balancing (i.e. gravity reduc
1 2, 12 1 2 . .2
V45 = ska (di + li -2dlicosé4) + sks (d5 + hi -2ds hAcosts)
(4d)
tion) are the following. From equation (8), the second term
disappears and the total potential energy equals Co which is
(4e) independent from any joint angle. When all the coefficients in
the second term are Zeros, the potential energy of the entire
system is constant. The static balancing conditions for the
10 system are:
The total potential energy of the apparatus, denoted by V, is
defined as the Sum of the gravitational and elastic potential C=0, i=1,2,..., 5 (9)
energy and is written as equation (5). Based on these conditions, a set of spring parameters is
selected as follows.
15
W = Wigm 1 + Vpigm2 + Vibody + Viru + (5)
Veg 1 + Veg2 + V,i + V2 + V3 + V.45 + V.67 k1 = (10)

Expressing the total potential energy in terms of the con

figuration variables, namely, the joint angles 0, (i-1,2,...,
7), equation (5) becomes equation (6), or alternately
expressed as equation (7).
25
7 (6)
W = Co -- X C;cost;
i=1

7 (7) Such springs preferably keep the total potential energy of

VTott = VMG + p VBG + Vs = Co -- X C;cost; 30 the system constant for any configuration and the gravity
i=1
force of the resulting system is preferably completely com
pensated.
The variables in equation (7) are defined as follows. A set of springs defined by equation (10) balances the
apparatus with a particular person where the preferred
35 springs stiffness values depend on mass m, which includes
Co = (8a) the human mass. In one embodiment, when the apparatus is
(2h1 + h)(m1 +m2)g + (hi-hs)(M + OMB)g + reneg +2h5m.4g + used by a different person, a new set of springs with different
stiffness values have to be installed, which is inconvenient. In
sks. ,(d;2,+ h;5)
.2
+ ks (d52 +. hi)
.2
+ k (di2,+ i)
2, 1 v.
+ 5X. k; (d.2,+ i)
12 another embodiment, the same set of springs is used for
i=1 40 everyone and the springs attachment points are preferably
C = 2n rig + l (2m2g + Mag + oMBG - kid) (8b) adjustable to change the values of d, (i-1, 2, . . . , 5) for
different persons. This latter embodiment is more preferable
C2 = 2m2 r2g + l2(MM g + piMBG-k2d2) (8c) in practice. Another embodiment of this adjustment of the
C3 = pn3rg - k3ds his (8d)
spring attachment locations comprises automation using
45 robotics technology
C4 = (2n4 + pint)rag + (n5 + pnu)lag - k4 dal (8e) Such a set of springs reduces the gravitational potential
energy of the human by a ratio p for all configurations. When
C5 = (n5 + ont)rsg - ksdsh4 (8f) p becomes one, it means that the gravity of the human
C6 = C4 (8g) attached to the apparatus is totally compensated, and there
50 fore, the human should experience a Zero-gravity condition.
C = Cs (8h) An adaptive reduced-gravity facility comprises a reduced
gravity simulation system. An adaptive reduced-gravity,
Mr. 4m.4 -- 2ns +ins (8i) human motion and dynamics experimental facility was used
MB = m3 + 2n + 2nt (8) to conduct researchonaerospace, bioengineering, and health
55 care. Computer-based simulation provides an analysis of the
reduced gravity simulation system. An instrumented nonhu
The term m is the total mass of the human and the harness man test investigates engineering principles and safety fea
and exoskeletons used to Support the human body and legs to tures. Ergonomics experiments use human Subjects to evalu
the balancing system. The value of m, varies from human to ate the reduced gravity system and assess any human factors
human. The joint angles 0s and 0 do not appear in equations 60 involved.
(8). These two angular displacements affect only the X and y FIG.3 is a schematic illustration of the adaptive reduced-G
location of the mass center of the device and have no effect on simulation facility. Mathematical modeling and simulation is
the Z location of the mass center. These two variable angles do used to enhance understanding of the physical world from
not affect the total potential energy of the system. Similarly, nanotechnology to astrophysics. The complexity of the
parametersh, handl do not influence the potential energy. 65 mechanics of human bodies requires math modeling and
The values of these parameters are arbitrarily chosen for other simulation in order to investigate biomedical applications. A
purposes in the design of the system. human body can be seen as an articulated multibody dynami
 US 8,152,699 B1
11
cal system with many degrees of freedom. Such a living
system is more difficult to model and analyze than other
physical systems like robots and vehicles.
The facility makes possible the measurement and under
standing of force-motion relations of human bodies. Cur
rently available human motion capture systems accurately
acquire kinematics data only, although some can capture lim
ited force data (e.g., the ground reaction forces) but the mea
Surement is insufficient for constructing accurate in-situ 10
human dynamics models.
The reduced-G facility provides a space to implement
methods to identify human dynamics parameters from the
simple exercise of a living human. Further, medical research
has shown that robotics devices can help improve neural 15 Example 2
rehabilitation and physiotherapy. The facility provides the
medical community with significant improvements to cur A non-limiting example of a system of an embodiment of
rently existing rehabilitation devices. the present inventions for the conditions given in equations
Existing facilities (e.g., neutral buoyancy lab, C-9 flight, (10) and (11) was based on the assumption that all the gravity
etc.) are primarily reserved for training astronauts because forces were balanced. The resulting apparatus was used for
they are complicated systems and expensive and difficult to simulating Zero-G scenarios. For planetary exploration, to
operate. The embodiment of the present invention provides an simulate a reduced-G Scenario as opposed to the Zero-G case,
innovative and low-cost alternative to current reduced-G there is a partial gravity balancing, done by replacing the
simulation technologies for the anticipated growing need of gravity acceleration term gin equations (10) and (11) with the
future manned space exploration missions to Moon and Mars. 25 term pg where p is a parameter that represents the ratio (or
The facility also comprises a motion simulator for training percentage) of the Earth gravity to be balanced.
space tourists for the projected 10,000-20,000 commercial Example 3
space travelers by 2020. There is significant benefit to having
training facilities available in which potential space travelers 30 A non-limiting example of a system to simulate walking on
train in reduced-G motion simulators prior to the flight. the Moon surface where the gravity is only /6 of the Earth
The facility comprises a safe and energy efficient passive gravity, has a corresponding pset to the following.
mechanical system; a seamless reduction of the gravity force
to any level from 0-g to 1-g; an active adaptation capability to
accommodate individuals’ different inertia distributions; 35 (12)
equipment to measure joint motions (angles and Velocities)
without occlusion; a 3D motion capture apparatus for both
free and constrained human activities; and multiple safety
protection measures for human safety. Example 4
The facility comprises the following research applications: 40
reduced-G physical simulation and training technology; A non-limiting example of a system for static balancing
human-body dynamics modeling and simulation; human-ma conditions used for a kinematics design of an apparatus was
chine interface and robotics; human anthropometric database statically balanced using springs. The following parameters
and computer animation; human-machine interface, human of the apparatus were known, i.e.,
factors and ergonomic optimization; and 3D human motion 45
analysis and performance optimization.
The facility comprises the following industrial applica m1 = m2 = 10 kg (13)
tions: training astronauts for EVA tasks on the Moon and m3 = 60 kg, m1 = 15 kg, ns = 10 kg; no = 20 kg
Mars; training space travelers for motion in reduced-gravity 50
i = i = 0.5 m; i = 0.35 m; is = 0.35 m.
environment; neural rehabilitation of people with walking h3 = 0.6 m; hd = 0.1 m.
impairment; human motion capture for computer animation, d = d2 = 0.4m; d = 0.6 m; d = 0.2 m; d5 = 0.15 m
movie and gaming industry; and entertainment for kids to r1 = 2 = 0.25 m; 3 = 0.5 m; r = 0.25 m;
experience a simulated 0-G or reduced-G feeling.
The dynamic performance of the reduced-gravity simula rs = 0.3 m; g = 9.81 m/s
55
tor, which is based on the technology of static balancing of
gravity force, was investigated as shown in the following The above mass values include the mass of a regular person
examples. attached to the apparatus. Substituting these parameters into
Equation (10), the results were:
Example 1 60

A non-limiting example of a system which vertically k = 4659.8 NFm; k = 4169.3 Nfm (14)
adjusts the spring attachment points A1, A2. . . . . and As (as k3 = 817.5 N/m; k = 1541.6 Nf m
illustrated in FIG. 2) based on the mass of the person on the
apparatus follows. The adjustment is based on the calcula 65 ks = 2943 Nf m
tions given in equations (11). It is preferable to mechanically
make a springs attachment points adjustable.
 US 8,152,699 B1
13 14
The subject human attached to the apparatus preferably felt simplify the analysis, the segments were modeled as rigid
no gravity force while he moved with the apparatus. The bodies, and the mass properties of the segments were defined
preferred described apparatus is based on kinematics and based on anthropometric databases and automatically set up
statics only, without considering any dynamics and friction by the Software, assuming a normal-fit person. The 18 mod
aspects of the system. The human subject felt the inertia and 5 eled joints, represented in FIG. 5, were upper neck 40, lower
friction forces when he/she interacted with the system. The neck 41, thoracic 42 and lumbar 43 on spinal, Scapular 44.
inertia forces/torques of the person were kept as is because shoulder 45, elbow 46, and wrist 47 on left and right arms, and
they were caused by the inertia property and dynamic motion hip 48, knee 49 and ankle 50 on left and right legs.
of the person, which was independent from the gravitation. In Although each joint contained three DOF and the model
an embodiment, the inertia forces/torques caused by the iner 10 provided a total of 54 DOF only 29 of them were active in the
tia of the apparatus itself were not present in a real reduced-G simulation. The rest of the DOF were locked either because
environment and thus were eliminated. Compensating an they did not physically exist (e.g. the frontal and transverse
inertia force is more difficult than compensating a gravity rotation of the knee), or were not necessary in the simulated
force and was done using an active control strategy. The activities for this study (e.g. some of the DOF in the neck do
inertia force of the apparatus is decreased when the apparatus 15 not move in normal walking). Some other DOF (e.g. the
comprises lightweight materials. frontal and transverse rotation of the lumbar, hip and ankle)
In another embodiment, joint friction torques were elimi were also eliminated, allowing focus on the motion in the
nated because they are not present if the apparatus is not Sagittal plane. Using this model and experimentally captured
attached to the person. A joint friction torque was decreased body-motion data in Cartesian space, the simulation Software
to a low level with proper mechanical design and proper was capable of computing the joint trajectories in joint space.
lubrication of the joint. Using these joint trajectories data, the joint torques that were
Using the above spring stiffness values and preferred required to produce the assumed body motion were calcu
dimensional and mass parameters, the gravity force of the lated.
system including the person was completely balanced. A Finally, the software applied the torques on the model
dynamic simulation model of the apparatus was implemented 25 joints and calculated the resulting ground reaction forces and
for verification using MSC ADAMSTM software, as shown in other reaction forces from interactions with the environment.
43. FIG. 4a illustrates a side view of the simulation results. The environment was modeled as a set of contact forces
FIG.4billustrates a front perspective view of the results of the between the human model and the ground or environment by
simulation results. The limitations of the software used to using the contact dynamics capability of MSC ADAMSTM.
model the results in FIG. 4 resulted in using fewer variables 30
Body motion data was obtained from either a motion capture
than are evident in FIG. 1. The simplified computer model experiment published in the literature or some other online
was used to verify the theory that Zero-gravity could be human motion database.
achieved. The dynamic simulation results illustrated in FIG. 4
shows that the apparatus remained at an arbitrarily assumed Example 6
initial configuration without moving in the gravity field. Such 35
a result indicated that the gravitational force was fully com The RGS model, as illustrated in FIG. 1, consisted of a
pensated, otherwise, the apparatus would be observed as inca 5-DOF platform and two 3-DOF leg exoskeletons. The plat
pable of staying at any given configuration without moving. form, including the two parallelograms sections, a strut and a
body harness assembly, was attached to the wall via a rota
Example 5 40 tional joint, which allowed the entire apparatus to rotate about
a vertical axis. The first and second parallelogram sections
Computer Simulation, Dynamic Simulation Model provided two additional DOFs, such that the body was able to
move forward and backward. The strut was connected to the
To fully verify the effectiveness of the reduced-gravity second parallelogram section through a rotational joint. The
simulator for astronaut training, hardware experiments with 45 upper body was fixed to a rigid plate which was part of the
human subjects should eventually be conducted. However, body-harness assembly. This rigid plate was connected to the
for safety consideration, Sufficient evidence regarding the Vertical strut through a hinge near the lumbar allowing the
feasibility and likelihood benefits of the new technology must human torso to turn forward (patching) during walking.
be first demonstrated by other means such as computer simu The 5-DOF platform provided the body a large workspace,
lations and nonhuman Subject testing before a human Subject 50 such that the body moved within the work space as freely as
test is planned. To this end, a simulation based feasibility possible. The leg exoskeletons, each of which included an
study of the invention was conducted. In the simulation study, upper parallelogram and a lower link, were capable of pivot
the dynamics behavior of the new system was quantitatively ing at the strut near the hip and also around the knee. The
assessed. upper parallelogram and the lower link were fixed to the
Due to the complexity and elasticity in its structure, the 55 upper and lower legs respectively. The platform, along with
human body was not easy to mathematically model accu the springs on it, was used to compensate the weight of the
rately. The commercial software LifeMODTM provided a whole body, while the leg exoskeletons and the springs were
powerful capability of constructing a dynamics model of a used to balance the weight of the legs and feet of the human.
human body. LifeMODTM is a plug-in toolbox for the multi The apparatus had a total of 13 independent rotational joints.
body dynamics simulator MSC ADAMSTM. With this soft 60
ware, a dynamics simulation model including a human model Example 7
and the gravity-reduction apparatus was implemented for the
study. Jumping Simulation
The human model, as illustrated in FIG. 5, was modeled as
a multibody system consisting of 19 rigid segments articu 65 A simulation was conducted to study the feasibility of the
lated by 18 joints. Each segment consisted of a group of proposed reduced-gravity simulator modeling the basic
human parts as they are located in the human anatomy. To human activity of jumping. The cases of free-body jumping
 US 8,152,699 B1
15 16
(referred to as “free jumping') was compared to the case of Example 8
jumping with the reduced-gravity apparatus (referred to as
“RGS jumping). In the simulation, it was assumed that the Walking
total mass of the human and the Suite was 200 kg.
A set of jumping kinematics data, available in CMU 5 A human model walking on a fixed flat floor was simulated
Graphic Lab Motion Capture Database, was used to generate using a set of ground walking data provided by LifeMODTM.
a jumping motion. The motions with four different reduction The data was captured from a ten-year-old child walking on a
ratios (i.e. X=0,0.5,0.62, and 5/6) were simulated. The ver flat floor. To imitate the relative motion between the human
tical displacement of the human mass center in “RGS jump 10
body and the treadmill, the first vertical bar of the RGS
ing” (jumping with the RGS device attached) and “free jump apparatus was specified to have a constant speed, rather than
ing” (jumping without the RGS device) with four different being fixed to the wall. This constant speed equaled the aver
reduction ratios were plotted in FIG. 6 and FIG. 7, respec age walking speed, such that the relative motion between the
tively. The peak height, peak time, and takeoff Velocity (i.e., human body and the floor was the same in average as if he/she
the velocity when leaving the ground) for both “free jumping 15 walked on a running treadmill and thus, the resulting system
and “RGS jumping situations in each of the four cases are response from the simulation was also valid.
listed and compared in Table 1. In the simulation, the springs were chosen to balance out%
of the Earth gravity force in order to simulate a moon envi
TABLE 1. ronment. The resulting ground reaction force and joint
Free Jumping vs. RGS Jumping torques were collected and compared with those of the same
person walking in the moon’s gravity field. Three materials
Peak Takeoff (i.e. aluminum, titanium and steel) were chosen as the struc
Height Peak Time Velocity tural materials of the apparatus to learn more about the influ
Case (mm) (sec.) (m/sec.)
ence of the mass of the apparatus. The same walking gait was
Free Jumping 25 used in the dynamics simulations of all the three cases.
p= 0 136.13 1.53 1.61 Example 9
(Earth)
p = 0.5 284.72 1.71 1.64
p = 0.62 34126 1.79 1.59 Comparison of Ground Reaction Force
(Mars) 30
p = 5/6 746.49 2.33 1.57
(Moon) The magnitude of the ground reaction forces on both the
RGS Jumping left and the right feet, which represents the interaction
between the human trainee and the environment, were plotted
p= 0 16342 1.58 1.45 in FIG.8. The time responses of the ground reaction forces in
(Earth) 35 all four cases are almost identical. These curves illustrated not
p = 0.5 333.81 1.88 1.45
p = 0.62 378.29 1.79 1.45 only a similar behavior but also close peak values. Thus, the
(Mars) resultant ground reaction forces which the human felt in the
p = 5/6 457.89 2.13 1.46 RGS walking were almost identical to that felt in the free
(Moon) walking on the Moon. This illustrated that the RGS technol
*All data were measured at body mass center 40 ogy is able to simulate a real reduced-gravity environment,
Such as that on the moon, at a reasonable level of accuracy.
FIG. 6 illustrates that the more the body weight was Simulation errors are list in Table 2. A simulation error
reduced, the higher the person’s jump and the longer the refers to the difference between the ground reaction forces of
person remained above the ground. However, compared with the RGS walking and that of the free walking on the Moon. As
the time history in the “free jumping (as shown in FIG. 7), 45 shown in Table 2, the absolute peak error is around 350N,
the time history of vertical displacement in “RGS jumping which is less than 18% of total body weight (i.e. 200 kg), and
was different, especially for a large reduction ratio (e.g. p=5/ the magnitude of the phase error is less than 0.1 sec., which is
6). This was mainly because when the apparatus was attached almost imperceptible to a human. The force errors were pri
to the human body, the motion of the human (e.g. jumping marily due to the inertia load of the RGS apparatus, which is
height in this case) was restricted by the workspace size of the 50 non-linear, but not proportional to the mass of the apparatus.
RGS apparatus.
Table 1 illustrates that the peak height in “RGS jumping Example 10
(except when p=5/6) was higher than that in “free jumping,
while the takeoff velocity was lower. The total kinetic energy Comparison of Joint Torques
of the human and the apparatus in “RGS jumping was 55
greater than that of the human alone in the “free jumping. The joint torques in the hip, knee and ankle joints in the
such that the resulting peak height in “RGS jumping, which sagittal plane, for both free walking and RGS walking were
corresponded to the potential energy, was also higher than plotted in FIG.9 for comparison. Joint torques in RGS walk
that in the “free jumping situation. The human jumped up ing were almost identical to those in the free-walking condi
with an apparatus. The human body and also the apparatus 60 tion. Joint torques represented the effort that the human made
gained kinetic energy as well as Velocity, and thus, it resulted in order to follow the proscribed walking gait. A larger torque
in a greater total kinetic energy, because the takeoff velocity required more effort of the human. The mass of the apparatus
of the human body in both situations was almost equivalent. also introduced additional inertia torques to the system, espe
The apparatus’s mass influenced the results. The apparatus cially in the hip and knee joints. However, since the human
provided a better simulation of the real reduced gravity con 65 body is an insensitive system (compared to instruments), it
dition when lightweight materials were used to construct the was possible to reduce the difference below the threshold by
RGS apparatus. using a lightweight apparatus.
 US 8,152,699 B1
17 18
TABLE 2 2. The reduced-gravity simulator assembly of claim 1
wherein said five degrees of freedom platform comprises at
The Largest Peak Error and Phase error least one parallelogram.
Material Left Foot Right Foot 3. The reduced-gravity simulator assembly of claim 1
wherein said five degrees of freedom platform comprises at
Largest Absolute/Relative Peak Error least one body harness.
(N7%)
4. The reduced-gravity simulator assembly of claim 1
Al (14.4 kg)* -296 -20.5% -15S -9.4% wherein said three degrees of freedom leg exoskeleton Sup
Ti (24 kg) -345 -24% -2.2 -0.1% ports and balances a weight of a human leg.
Steel -182 12.6% -61 -3.7% 10
(41.6 kg) 5. The reduced-gravity simulator assembly of claim 1 com
Largest Phase Error (ms) prising two leg exoskeletons comprising at least three degrees
of freedom.
Al (14.4 kg) -40 50
Ti (24 kg) -40 8O 6. The reduced-gravity simulator assembly of claim 1
Steel -20 60 15 wherein an upper portion of said leg exoskeleton attaches to a
(41.6 kg) human thigh and a lower portion of said leg exoskeleton
*The number in the brackets represents the approximate weight of the RGS apparatus attaches to a human lower leg.
associating with th material,
gravity-balanced multi-DOF passive apparatus was used 7. The reduced-gravity simulator assembly of claim 1
to simulate walking and jumping in a reduced-gravity envi wherein said five degrees of freedom platform Supports an
ronment, thus providing an alternative simulation technology entire human body weight.
for training astronauts for planetary exploration missions. 8. The reduced-gravity simulator assembly of claim 1
The reduced-gravity simulator, along with the human, was wherein one side of said backplate is disposed adjacent to a
modeled as a multi-body dynamical system. human torso.
Two activities, namely, jumping and walking, were simu 9. The reduced-gravity simulator of claim 1 further com
lated and compared between the free-body condition (i.e. 25 prising a plurality of rotating couplings comprising hinges.
without the reduced-gravity simulator) and the constrained 10. The reduced-gravity simulator of claim 5 further com
body condition (i.e. with the reduced-gravity simulator). The prising eleven independent degrees of freedom.
simulation results showed that a human in the RGS-con 11. The reduced-gravity simulator of claim 1 further com
strained case tended to produce a similar dynamic behavior to prising a plurality of connectors comprising spring attach
the one in the free body case. Also, the ground reaction forces 30 ment points.
and joint torques that a human is Subject to in RGS walking 12. A method of initializing a reduced-gravity simulator
was almost identical to that in the real reduced-gravity envi comprising:
ronment. The proposed RGS apparatus was capable of simu attaching the reduced-gravity simulator assembly to a user;
lating human walking in a realistic reduced-gravity condition.
Additional inertia and friction loads due to the mass and attaching at least one leg exoskeleton to a lower portion of
joints of the RGS apparatus were applied to the human body 35 a strut and a leg of the user, the exoskeleton comprising
when the human was exercising with the apparatus. These at least three degrees of freedom;
additional loads were reduced by using lightweight materials attaching at least one spring to a spring attachment point on
and proper design of the apparatus. the exoskeleton;
The preceding examples are repeatable with similar Suc sliding the spring attachment point relative to a portion of
cess by Substituting the generically or specifically described 40 the exoskeleton until the pressure underneath the user is
initial conditions and/or operating conditions of this inven at least partially eliminated;
tion for those used in the preceding examples. maintaining a constant potential energy of the assembly by
Although the invention has been described in detail with adjusting the spring attachment point location to stati
particular reference to these preferred embodiments, other cally balance the gravity force at a working configura
embodiments can achieve the same results. Variations and 45 tion; and
modifications of the present invention will be obvious to those adjusting the stiffness values of the spring Such that the
skilled in the art and it is intended to cover all such modifi user feels as if they are in a reduced gravity environment.
cations and equivalents. The entire disclosures of all refer 13. The method of claim 12 further comprising using a
ences, applications, patents, and publications cited above treadmill to allow continuous forward walking.
and/or in the attachments, and of the corresponding applica 50 14. The method of claim 12 wherein attaching to the user
tion(s), are hereby incorporated herein by reference. comprises:
What is claimed is: attaching the users torso to a backplate;
1. A reduced-gravity simulator assembly comprising: allowing the torso to freely move in five degrees of free
a platform comprising at least five degrees of freedom, said dom;
platform comprising a strut and at least one spring, a first 55 Supporting at least one of the user's legs by the three
end of said spring attached to an upper portion of said degrees of freedom exoskeleton allowing free move
Strut, and ment of the leg; and
a second end of said spring attached to a first end of a providing redundant degrees of freedom about the Z axis to
backplate;
a second end of said backplate attached directly to a lower allow ergonomic comfort while the user is walking.
portion of said strut and hingeably moveable with 60 15. The method of claim 12 wherein the step of sliding the
respect to said lower portion of said strut, spring attachment point relative to a portion of the exoskel
at least one leg exoskeleton comprising at least three eton comprises:
degrees of freedom attached to said lower portion of said sliding the spring attachment points and measuring pres
Strut, said leg exoskeleton comprising a spring attach Sure until the measured pressure underneath the users
ment point which is slideable with respect to a portion of 65 foot is completely gone but the foot still touches the
said leg exoskeleton; and treadmill;
a treadmill. changing the body and/or leg pose; and
 US 8,152,699 B1
19 20
repeatedly sliding the spring attachment point until 19. The method of claim 12 further comprising the steps of:
required balancing is achieved. Scaling the spring attachment point based on the percent
16. The method of claim 15 wherein measuring pressure is age of gravity which is to be reduced; and
measured manually. retaining a percentage of the pressure under a foot of the
17. The method of claim 15 wherein measuring pressure is user because partial gravity is still present.
measured automatically.
18. The method of claim 17 wherein measuring pressure
automatically comprises installing an actuated spring adjust
ment device for the spring.