Human-robot haptic joint actions is an equal control-sharing approach possible?

This paper proposes a speculative model and control framework for human-robot haptic joint (i.e. collaborative) actions in which the roles human and robot partners play as leaders or followers is not predetermined prior to tasks execution. In this proposed scheme, role switching is made by a homotopy: that is, a time-variant, continuous or set-value, weighting interpolation function between the two extreme (leader of follower) behaviors for each individual. The role of each individual is changeable anytime with any plausible dynamic all along task execution. The formalism allows arbitrary functions for each partner's leader and follower control laws. We discuss strengths, weaknesses and perspectives of such an approach.