TENYEARSIN
LOCOMOTIONCONTROLRESEARCH
JeheeLee
SeoulNationalUniversity
[SIGGRAPH 2010] Lee et al, Data-driven biped control
[SIGGRAPH 2010] Lee et al, Data-driven biped control
[SIGGRAPH 2010] Lee et al, Data-driven biped control
Hubo
Before 2007
After 2007
Simplified dynamics Model
Fullbody dynamics
Feedback only
(stereotyped robotic walking)
Analytic balance strategy
Derivative-based optimization
(conjugate-gradient, Newton, BFGS, …)
Feedback and feedforward
(motion capture references)
Learning from experience
Derivative-free optimization
(CMA-ES)
Before 2007
After 2007
Simplified dynamics Model
Fullbody dynamics
(inverted pendulum)
Feedback only
(stereotyped robotic walking)
Analytic balance strategy
Feedback and feedforward
(motion capture references)
Computational model of balancing
(regression, learning from experience,
optimization at runtime)
Derivative-based optimization
(conjugate-gradient, Newton, BFGS, …)
Derivative-free optimization
(CMA-ES)
[SIGGRAPH 2007]
Sok et al, Simulating Biped Behaviors from Human Motion Data
[SIGGRAPH 2010] Lee et al, Data-driven biped control
Before 2007
After 2007
Simplified dynamics Model
Fullbody dynamics
(inverted pendulum)
Feedback only
(stereotyped robotic walking)
Analytic balance strategy
Feedback and feedforward
(motion capture references)
Computational model of balancing
(regression, learning from experience,
optimization at runtime)
Derivative-based optimization
(conjugate-gradient, Newton, BFGS, …)
Derivative-free optimization
(CMA-ES)
PlausibilityofSimulation
PhysicalCorrectness
ThesimulationiscorrectwithrespecttoNewton’slawofmotion
Nofictionalforceappliestothebody
AdmissibleControl
Controlforce/torquesarevalidwithinmusclecapacity
GRF(groundreactionforce)consistentwithcontrolforce/torque
PlausibilityofSimulation
TypeI(Stronglyadmissible)
Simulationisphysicallycorrectandcontrolisadmissible
TypeII(weaklyadmissible)
Simulationisphysicallycorrect,butcontrolmaynotbeadmissible
ex)GRFsarecomputedasoptimizationparametersindependentofjointtorques
TypeIII(Visuallyplausible)
Physicalcorrectnessisnotguaranteed
ex)Fictionalforcemayapplyatcontactpoints
Dynamics
Energertic
Agility
Stability
Skin
Low-energy
Balance
Muscle
Static
Pertubation
Robustness
Applications
Humanoid
Robot
Emotion
Fatigue
Aging
Quadruped
Video
Games
Adapation
Group
Biped
Social
Flying
High-Level Behavior
Modeling
Skeleton
Biological
Motion
Simulation
Gait
Analysis
Tendon
Interaction
Type
Dynamics
Energertic
Agility
Stability
Skin
Low-energy
Balance
Muscle
Static
Pertubation
Robustness
Applications
Humanoid
Robot
Emotion
Fatigue
Aging
Quadruped
Video
Games
Adapation
Group
Biped
Social
Flying
High-Level Behavior
Modeling
Skeleton
Biological
Motion
Simulation
Gait
Analysis
Tendon
Interaction
Type
L
Gait2562
Gait2592
Fullbody
(25DOFs,62muscles) (25DOFs,92muscles) (39DOFs,120muscles)
[SIGGRAPH Asia 2014] Lee et al, Many-Muscle Humanoids
18
[SIGGRAPH Asia 2014] Lee et al, Many-Muscle Humanoids
Dynamics
Energertic
Agility
Stability
Skin
Low-energy
Balance
Muscle
Static
Pertubation
Robustness
Applications
Humanoid
Robot
Emotion
Fatigue
Aging
Quadruped
Video
Games
Adapation
Group
Biped
Social
Flying
High-Level Behavior
Modeling
Skeleton
Biological
Motion
Simulation
Gait
Analysis
Tendon
Interaction
Type
UnilateralPainfulAnklePlantarFlexor
Patientstendtoreducethe
useoftheankleplantarflexors
PainfulJointsonUnilateralLimb
Patientstendtoreducecontactforce
PainfulLeftAnklePlantarFlexor
PainfulJointsonLeftLeg
WaddlingGait
BilateralGluteusMedius&
MinimusWeakness
Upperbodyswinglaterally
TrendelenburgGait
UnilateralGluteusMedius&
MinimusWeakness
Dynamics
Energertic
Agility
Stability
Skin
Low-energy
Balance
Muscle
Static
Pertubation
Robustness
Applications
Humanoid
Robot
Emotion
Fatigue
Aging
Quadruped
Video
Games
Adapation
Group
Biped
Social
Flying
High-Level Behavior
Modeling
Skeleton
Biological
Motion
Simulation
Gait
Analysis
Tendon
Interaction
Type
Balance and Stability
Under what conditions is human gait more stable?
What factors affect the level of stability?
Are simulated walking as stable as human walking?
Do the factors that affect human gait also influence
controller stability?
[SIGGRAPH Asia 2015] Lee et al, Push-Recovery Stability
Four factors that affect gait stability
Level of crouch
Walking speed
Magnitude of push
Timing of push
Crouch Gait is more stable than Normal Gait
It detours less if
it walks faster,
push is weaker, and
push happens later in the swing phase
Similar trends for humans and simulation
Applications in Clinical Gait Analysis
Surgery improves cerebral palsy gait
by lengthening tight muscles/tendons and fixing bone deformity
Predictive simulation of post-operative gaits
from pre-operative motion capture and surgery planning
Dynamics
Energertic
Agility
Stability
Skin
Low-energy
Balance
Muscle
Static
Pertubation
Robustness
Applications
Humanoid
Robot
Emotion
Fatigue
Aging
Quadruped
Video
Games
Adapation
Group
Biped
Social
Flying
High-Level Behavior
Modeling
Skeleton
Biological
Motion
Simulation
Gait
Analysis
Tendon
Interaction
Type
Papers & Videos are available at
http://mrl.snu.ac.kr/