Adaptive multi-objective control explains how humans make lateral maneuvers while walking.

Area of Science:

• Biomechanics
• Human Motor Control
• Robotics

Background:

• Humans maintain balance during locomotion by modulating foot placement.
• Lateral stability is a key challenge in human walking.
• Previous models described lateral stepping during straight-ahead walking using Goal Equivalent Manifolds (GEMs).

Purpose of the Study:

• To investigate if humans regulate lateral foot placement during non-steady-state maneuvers.
• To test if error minimization consistent with goal functions explains lateral stepping during transient tasks.
• To extend the existing framework for lateral stepping regulation to dynamic maneuvers.

Main Methods:

• Twenty healthy young adults performed lateral lane-change maneuvers in a virtual reality environment.
• A theoretical framework based on GEMs was adapted for transient walking tasks.
• Computational models were used to analyze step-to-step adaptation of stepping behavior.

Main Results:

• Human stepping behavior during lateral maneuvers was consistent with theoretical predictions.
• Models incorporating step-to-step parameter adaptation accurately predicted observed behavior.
• The study demonstrated rapid, real-time adaptation of motor control strategies.

Conclusions:

• Humans can regulate lateral foot placement during complex, non-steady-state maneuvers.
• Evolving cost landscapes are likely used in real-time for adaptive motor tasks.
• The findings extend the predictive power of stepping regulation frameworks to diverse walking scenarios.