Abstract

Co-manipulation is when two or more people work together to move an object. In this work, we consider the physical co-manipulation of objects that are large and heavy enough that they are more easily handled by two or more participants. We currently lack a fundamental understanding of how people perform such tasks together, both the basic signals they use to communicate and how those signals are mapped to actions and behaviors. This work builds upon the existing body of knowledge to improve our understanding of physical co-manipulation. Specifically, interaction forces are often considered a key method for information transfer in co-manipulation, but the term is broadly and variably defined. We select a definition and introduce the net force method, which decomposes interaction force into forces contributing to the magnitude of acceleration and those that are canceled by other forces acting on the object. This method is suitable for multi-agent (with two or more agents) co-manipulation in 6 degrees of freedom with large, heavy objects, and can be implemented using a robotic system with a force-torque sensor and an accelerometer. We also present two other decomposition methods that meet these criteria but are less applicable to our specific goals. Extracting consistent patterns and behaviors from existing co-manipulation studies proved challenging. To address this, an experiment was designed to analyze basic motions in all six degrees of freedom and how they extend to two and three degrees of freedom. Data was examined for regular patterns, noise levels in the various signals during rest states, and identifiable transitions from static to active states. Analysis showed that velocity is the ideal measure for predicting motion, except at the beginning or end of movement when velocity is near zero. The study also found that all but one of the six degrees of freedom are communicable via a sense of force and motion, or the haptic channel. Additionally, it revealed that for planar tasks requiring two degrees of freedom, participants tend towards minimum effort trajectories; however, this pattern does not apply to tasks in the vertical plane.

Degree

MS

College and Department

Ira A. Fulton College of Engineering; Mechanical Engineering

Rights

https://lib.byu.edu/about/copyright/

Date Submitted

2024-08-12

Document Type

Thesis

Keywords

co-manipulation, interaction force, haptics

Language

english

Download

Included in

Engineering Commons

Share

COinS