A Kinematic Model to Predict a Continuous Range of Human-Like Walking Speed Transitions

Area of Science:

• Biomechanics
• Human locomotion
• Gait analysis

Background:

• Constant speed walking is understood, but speed transitions are not.
• It is unclear if gait transitions are smooth or involve unique steps.

Purpose of the Study:

• To investigate how humans change walking speed.
• To determine if gait transitions smoothly morph or consist of unique steps.

Main Methods:

• Analyzed joint kinematic data from treadmill walking speed transitions.
• Decomposed gait into trend and periodic components.
• Developed and validated a predictive kinematic model using Bezier polynomials and Fourier series.

Main Results:

• Gait transitions were 2-3 steps long and began near the speed change.
• A predictive kinematic model accurately predicted untrained gaits regardless of training data.
• Similar prediction errors across training sets suggest smooth morphing of joint kinematics.

Conclusions:

• Human joint kinematics smoothly morph during walking speed changes.
• The transition between different walking speeds is a continuous process.