Expressive Robot Motion Timing

Abstract

Our goal is to enable robots to time their motion in a way that is purposefully expressive of their internal states, making them more transparent to people. We start by investigating what types of states motion timing is capable of expressing, focusing on robot manipulation and keeping the path constant while systematically varying the timing. We find that users naturally pick up on certain properties of the robot (like confidence), of the motion (like naturalness), or of the task (like the weight of the object that the robot is carrying). We then conduct a hypothesis-driven experiment to tease out the directions and magnitudes of these effects, and use our findings to develop candidate mathematical models for how users make these inferences from the timing. We find a strong correlation between the models and real user data, suggesting that robots can leverage these models to autonomously optimize the timing of their motion to be expressive.

Differences of Human Perceptions of a Robot Moving using Linear or Slow in, Slow out Velocity Profiles When Performing a Cleaning Task

“…For example, a video study used the principles of Anticipation and Follow Through and Overlapping Action to make it easier for viewers to understand what a non-humanoid robot was intending to do [11]. Another study used the timing principle to move a robot's arm, which helped people attribute meaning to the motion [12].…”

Animation Techniques in Human-Robot Interaction User Studies

“…Several HRI studies examined how individuals ascribe feelings and personalities to the way robots move, whether they look like a dog [11,8], a vacuum cleaner [30,79,73], or simply an arm [100]. Other HRI animacy studies are based on Piaget and examine children's relationship to robots and other things that are alive [56,61,14].…”

My Parents

“…Nikolaidis et al [2017a] have contributed action planning algorithms that allow their robot to reveal its capabilities adaptively through a game theoretic model of human expectations. Sciu i et al [2014], Zhou et al [2017] have developed expressive robotic lifting motions to help humans understand the weights of the objects that robots are manipulating. The ease with which a person could recognize a robot's goals by observing its action execution improves robot legibility [Dragan et al, 2013, Dragan and Srinivasa, 2014], predictability [Fisac et al, 2016, acceptance [Cha et al, 2015], and naturalness [Szafir et al, 2014], which are important for human recognition of robot tasks and human-robot collaboration [Powers andKiesler, 2006, Gielniak et al, 2013].…”

“…One approach to resolve this issue is to develop transparency in robot behaviors by enabling a robot to provide comprehensible explanations [Seegebarth et al, 2012, Zhou et al, 2017. There are multiple media in which explanations could be transferred from robots to humans.…”

“…People have focused on using robot motion as a demonstration to expressively convey robot capabilities [Nikolaidis et al, 2017a], robot learning progresses [Nicolescu and Mataric, 2003], object physical properties [Sciu i et al, 2014, Zhou et al, 2017, and robot goals [Gielniak et al, 2013, Dragan and Srinivasa, 2014, Szafir et al, 2014, Takayama et al, 2011. Zhang et al [2015] introduce plan explicability as the ease with which humans associate tasks with robot actions in the plan and predictability as the ease with which humans to predict the next task given actions in the previous tasks.…”