Fast Handovers with a Robot Character: Small Sensorimotor Delays Improve Perceived Qualities

Pan, Matthew K. X. J.; Knoop, Espen; Bächer, Moritz; Niemeyer, G.

doi:10.1109/iros40897.2019.8967614

Cited by 24 publications

(24 citation statements)

References 20 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Here, improvement of the interaction with particular human users would not only result directly from the representation of inter-individual differences, but also from a general approximation of human behavior. First, human-like, e.g., less precise but more versatile, robot movements have been shown to improve the perceived interaction quality (Pan et al, 2019). Second, cognitive modeling has been successfully used for pretraining machine learning models (Bourgin et al, 2019), which increases learning efficiency and has strong potential to foster distinct progress in personalizing interaction.…”

Section: Resultsmentioning

confidence: 99%

Personalizing Human-Agent Interaction Through Cognitive Models

Schürmann

Beckerle

2020

Front. Psychol.

View full text Add to dashboard Cite

Cognitive modeling of human behavior has advanced the understanding of underlying processes in several domains of psychology and cognitive science. In this article, we outline how we expect cognitive modeling to improve comprehension of individual cognitive processes in human-agent interaction and, particularly, human-robot interaction (HRI). We argue that cognitive models offer advantages compared to data-analytical models, specifically for research questions with expressed interest in theories of cognitive functions. However, the implementation of cognitive models is arguably more complex than common statistical procedures. Additionally, cognitive modeling paradigms typically have an explicit commitment to an underlying computational theory. We propose a conceptual framework for designing cognitive models that aims to identify whether the use of cognitive modeling is applicable to a given research question. The framework consists of five external and internal aspects related to the modeling process: research question, level of analysis, modeling paradigms, computational properties, and iterative model development. In addition to deriving our framework from a concise literature analysis, we discuss challenges and potentials of cognitive modeling. We expect cognitive models to leverage personalized human behavior prediction, agent behavior generation, and interaction pretraining as well as adaptation, which we outline with application examples from personalized HRI.

show abstract

Section: Resultsmentioning

confidence: 99%

Personalizing Human-Agent Interaction Through Cognitive Models

Schürmann

Beckerle

2020

Front. Psychol.

View full text Add to dashboard Cite

show abstract

“…As discussed, one of the main advantages of PPMPs is to be a semi-model-free approach which only requires the general dynamics of coupled oscillators to generate predictive adaptation. In contrast, take for example the recent work of Pan et al [24] who implemented a handover controller where the joint trajectories were hard-coded for each degree of freedom of a 7-DoF arm. To account for the case where the robot needed to retract its arm, Pan et al used a kind of finite-state machine approach to activate the action for arm retraction.…”

Section: Ppmp Vs Hard-coded Trajectoriesmentioning

confidence: 99%

“…In PPMPs, the arm retraction is achieved without additional high-level rules, as it is simply the result of the adaptation provided by the coupled oscillators. Also, PPMPs dynamically control the timing along the entire trajectory while in [24] only the initial delay can be manually adjusted. The reader is invited to watch the video https://www.youtube.com/watch?v=w1Ff4nqcUvk of Pan et al, and compare with the accompanying handover video of this paper https://gjmaeda.github.io/ videos/PPMP/PPMP_handover.mp4 while qualitatively observing the similarities in the robot's reaction.…”

Section: Ppmp Vs Hard-coded Trajectoriesmentioning

confidence: 99%

Phase portraits as movement primitives for fast humanoid robot control

Maeda¹,

Koc

Morimoto³

2020

Neural Networks

View full text Add to dashboard Cite

Currently, usual approaches for fast robot control are largely reliant on solving online optimal control problems. Such methods are known to be computationally intensive and sensitive to model accuracy. On the other hand, animals plan complex motor actions not only fast but seemingly with little effort even on unseen tasks. This natural sense of time and coordination motivates us to approach robot control from a motor skill learning perspective to design fast and computationally light controllers that can be learned autonomously by the robot under mild modeling assumptions. This article introduces Phase Portrait Movement Primitives (PPMP), a primitive that predicts dynamics on a low dimensional phase space which in turn is used to govern the high dimensional kinematics of the task. The stark difference with other primitive formulations is a built-in mechanism for phase prediction in the form of coupled oscillators that replaces model-based state estimators such as Kalman filters. The policy is trained by optimizing the parameters of the oscillators whose output is connected to a kinematic distribution in the form of a phase portrait. The drastic reduction in dimensionality allows us to efficiently train and execute PPMPs on a real human-sized, dual-arm humanoid upper body on a task involving 20 degrees-of-freedom. We demonstrate PPMPs in interactions requiring fast reactions times while generating anticipative pose adaptation in both discrete and cyclic tasks.

show abstract

“…Different aspects of human–robot hand‐over have been studied, e.g., ergonomy and psychological effects on the human worker [14–26], motion‐planning [20–40], and grasping [40–44].…”

Section: Introductionmentioning

confidence: 99%

“…Unfortunately, only the human‐to‐robot case is considered and if the operator suddenly moves the hand, the robot must go back to the home configuration and replan the trajectory. In Pan et al [37], passive markers are placed on the object that must be exchange to track its pose and use this information as input of the control algorithm. The movements of the human operator are not acquired, so the control hand‐over algorithm is not able to modify online the trajectory of the robot during robot‐to‐human hand‐over.…”

Section: Introductionmentioning

confidence: 99%

Vision‐based control architecture for human–robot hand‐over applications

Melchiorre

Scimmi

Mauro

et al. 2020

Asian Journal of Control

View full text Add to dashboard Cite

Human-robot collaboration will be an essential part of the production processes in the factories of tomorrow. In this paper, a human-robot handover control strategy is presented. Human and robot can be both giver and receiver. A path-planning algorithm drives the robotic manipulator towards the hand of the human and permits to adapt the pose of the tool center point of the robot to the pose of the hand of the human worker. The movements of the operator are acquired with a multi 3D-sensors setup so to avoid any possible occlusion related to the presence of the robot or other dynamic obstacles. Estimation of the predicted position of the hand is performed to reduce the waiting time of the operator during the handover task. The hardware setup is described, and the results of experimental tests, conducted to verify the effectiveness of the control strategy, are presented and discussed.

show abstract

Fast Handovers with a Robot Character: Small Sensorimotor Delays Improve Perceived Qualities

Cited by 24 publications

References 20 publications

Personalizing Human-Agent Interaction Through Cognitive Models

Personalizing Human-Agent Interaction Through Cognitive Models

Phase portraits as movement primitives for fast humanoid robot control

Vision‐based control architecture for human–robot hand‐over applications

Contact Info

Product

Resources

About