Preparing to move: Setting initial conditions to simplify interactions with complex objects.

Area of Science:

• Human motor control
• Robotics
• Human-computer interaction

Background:

• Human interaction with objects involves complex dynamics, especially with liquids.
• Object dynamics can be nonlinear and unpredictable, depending on initial states.
• Predictable object-hand interactions may reduce online error correction and aid feedforward control.

Purpose of the Study:

• To investigate how humans explore and prepare object dynamics for task execution.
• To test the hypothesis that subjects seek initial conditions that shorten transients for stable, predictable interactions.
• To evaluate alternative hypotheses: reduced effort, increased smoothness, and reduced risk.

Main Methods:

• A simplified cup-and-ball model was used in a virtual environment.
• Human subjects interacted with the virtual object using a force-feedback robotic manipulandum.
• Subjects explored and prepared the object before performing a rhythmic movement task.

Main Results:

• Subjects increased interaction force predictability and converged on initial conditions that decreased transients.
• The primary hypothesis was supported; alternative hypotheses (effort, smoothness, risk) were not.
• Subject strategy was more effortful and less smooth than in simple reaching movements.

Conclusions:

• Humans actively prepare objects to minimize transients and enhance stability for predictable interactions.
• Chosen initial conditions matched simulations predicting the most predictable interactions.
• Findings support the hypothesis that predictability, not just smoothness or effort, guides motor preparation.