Animating Human Athletics

1
Animating Human Athletics

• Jessica K. Hodgins, Wayne L. Wooten,
• David C. Brogan and James F. OBrien
• SIGGRAPH 1995
• Jaeho Kim, VR Lab.

2
Animating Human Athletics

• Dynamic simulation of human motion
• Running
• Cycling
• Vaulting

3
Preliminary

• Passive system
• Active system
• Torque
• Inverse kinematics

4
Passive systems vs. Active systems

• Passive systems
• Only external forces alter the motion of the
  system
• Examples Leaves, water spray, and clothing
• Active systems
• Motors, muscles, or other internal forces control
  the motion of the system
• Examples Running human, and swimming fish

5
Torque

• Rigid body dynamics
• Newtons equation for translations
• Eulers equation for rotational motion
• Torque

6
Inverse kinematics

• Inverse kinematics
• Given location and orientation of end effector,
  find joint variables

?
7
Animating Human Athletics

• Dynamic simulation of human motion
• Running
• Cycling
• Vaulting
• Control algorithms
• state machines that describe each specific motion

8
Related work

• Robotics
• Biomechanics
• Motion capture data
• Muscle activation data
• Energy curve
• Stance duration, flight duration, step length
• Computer graphics
• Jack system

9
Problems
10
Problems

• Human models ?
• Dynamics ?
• Controls ?
• Visualization ?

System design
11
System design

• Active system

desired behavior
Forces and torques
Model
User
Control
Numerical Integrator
state
graphics
12
System design

• Active system

desired behavior
Forces and torques
Model
User
Control
SDFast from Symbolic Dynamics
state
graphics
13
Human model

• Between 15 and 17 regid bodies connected with
  rotary joints, allowing 22 to 32 controlled DOF
• Body segment densities obtained from
  biomechanical data
• Mass and moments of inertia calculated from the
  polygonal model

14
Control Algorithms
Forces and torques
desired behavior
Control

• state machines connecting phase of behavior to
  active control laws

15
Control Algorithms
Forces and torques
desired behavior
Control

• Control the primary actions using equations for
  motion
• Basic process (for each time step)
• calculate joint positions and velocities
• compute joint torque (with proportional-derivative
  servos)
• integrate equations of motion
• Hand designed and tuned

16
Control Laws
Forces and torques
desired behavior
Control

• Proportional-derivative servos
• To compute the force required to move joints to
  the desired position

17
System design

• Active system

desired behavior
Forces and torques
Model
User
Control
SDFast from Symbolic Dynamics
state
graphics
18
Human Motion

• Running
• Bicycling
• Vaulting

19
Running
20
Active Leg

• At touchdown
• The desired distance from the hip to the heel
  projected onto the ground plane
• To compute the desired knee and hip angle
• ? xhh, yhh, and zhh,
• ? the inverse kinematics of the leg
• Related to forward velocity

21
(No Transcript)
22

• Proportional-derivative servos are used to
  compute torques for the hip joint of the stance
  leg

23
Idle Leg

• To be shortened so that the toe does not stub the
  ground
• The hip angles mirror the motion of the active leg

24
Sh0ulder
25
(No Transcript)
26
Bicycling

• (two-sided) Spring and damper systems
• The hands to the handlebars
• The feet to the pedals
• The crank to the rear wheel

27
The desired torque at the crank

• The desired torque at the crank
• The desired forces from the left and right legs

28
Steering

• a roll angle
• ß yaw angle

29
(No Transcript)
30
Vaulting

• Spring board
• Spring and damper model

31
Putting the hands on the horse
32
Balance controller

• Center of mass must remain over the support
  polygon
• Controller derives desired joint angles based on
  the error between the desired center of mass and
  the actual center of mass
• Knee angle computed to maintain body height

33
(No Transcript)
34
Higher-level behaviors

• Choreographing an animation with many bicyclists
  or runners
• difficult !!!
• Implemented the Reynolds algorithm
• To control a group where the members have
  dynamics
• Refer to Brogan and Hodgins, Group behaviors for
  systems with significant dynamics, IEEE/RSJ
  International Conference on Intelligent robot and
  systems, 1995

35
Secondary motions

• Sweatpants
• Splashing water

36
(No Transcript)
37
Conclusion

• Dynamic simulation of human motion
• Running
• Cycling
• Vaulting
• Control algorithms
• state machines that describe each specific motion
• Control the primary actions using equations for
  motion

38
Discussion

• Advantages
• produce physically correct/realistic motions
• easy to create simulated motion
• can easily create similar motions
• Disadvantages
• robust algorithms difficult to create
• require detailed knowledge of the system
• computational expense grows with constraints
• generally accurate only for one complete action

Human simulation, Keith Thoresz, Suan Yong,
April 6, 1999
39
References

• Marc H. Raibert, Jessica K. Hodgins, Animation
  of Dynamic Legged Locomotion, SIGGRAPH 1991.
• Jessica K. Hodgins, Three-Dimensional Human
  Running, In Proceedings of IEEE Conference on
  Robotics and Automation, 1996.
• Dynamic Behaviors for Real-Time Synthesis
  Humans, SIGGRAPH 1995 Course Notes