PHYSICAL HUMAN-MACHINE COORDINATION

Problem + Background

Imagine the world 100 years from now, robots are everywhere walking, jumping and rolling among humans in society. However, they have not taken over the world, as many may ironically fear, but they have seamlessly become a part of our physical world. Physically coordinating with humans to move furniture or assist in physical rehabilitation or handle fragile material with smooth and intuitive adjustments that humans may feel but have difficultly explaining. These types of joint action require physical human-machine coordination and highlight the importance of sensorimotor exchanges and its influence on the physical interaction as a whole, as well as the actions and behaviors of the individuals that make the whole. As robots are rapidly becoming a major part of our physical world, especially in diverse work environments involving physical joint action, it is important to consider how these subtle, yet significant sensorimotor exchanges shape our attitude towards the machine and how our attitude may be represented through sensorimotor exchange. Such attitudes that can influence these interactions and could possibly be measurable through sensorimotor exchanges, is our physiological construct of trust as it is translated through our physical reliance on the machine. However, our understanding of these exchanges are difficult to acknowledge and interpret through typical methods of analysis and inferential statistics.

Process

In an interdisciplinary collaborative effort, the ASU ADAPT lab (Automation Design Advancing People and Technology) lab and ASU RISE (Robotics and Intelligent SysEms) developed and designed experiments to examine sensorimotor exchanges and constructs of trust in physical human-machine coordination. The experiments primarily focused on coordination in physical joint tasks between human and human-robot dyads. Our first experiment we examined the relationship between psychological construct of trust, joint physical coordination, and reliance on the machine between human-robot dyads performing a physical joint action task. Our other experiment examined the same constructs in a similar physical joint action task but with a surprise occurring in one of the tasks, and through human dyads. The human dyads experiment was primarily examining various definitions of trust between and within dyads and how trust may evolve throughout the interaction when a surprise occurs. We were also examining the sensorimotor exchanges through the dyads force, velocity, and trajectory with 3D motion sensors.

Objectives

Examine the relationship between reliance on a machine and trust as a physiological construct.
Explore sensorimotor exchanges between human dyads and identify patterns.
Interpret patterns of sensorimotor exchanges with dyads force data input and analyze using cross recurrence quantification analysis (CRQA).

Outcome

In our first study with human-robot dyads our finding suggests that physical reliance on a machine, through interaction force, relates to trust but interaction force is only one aspect of the relationship. Many other factors need to be studied to build a more diverse account of trust and reliance in physical human-machine coordination that considers more realistic human-robot interactions. Considering these kinds of real-world human-robot interactions could help develop robot control algorithms that could better predict human intention and behavior and create smoother and more intuitive adjustments in physical joint action.

While examining sensorimotor exchanges, specifically force and motion signals, within human dyads performing physical joint action tasks, I am exploring an untraditional method of analysis called cross recurrence quantification analysis (CRQA). CRQA is a nonlinear correlational analysis technique that assesses correlation (or coupling) between two time-series and assesses coupling through the repetition of elements or patterns in sequence. With CRQA I am trying to understand how human dyads exchange force and motion signals and identify patterns or recurrence within human dyads force and motion signals overtime. More specifically I am examining these sensorimotor signal exchanges in situations of uncertainty to understand how these signals are used to haptically communicate, coordinate and adapt. CRQA can help us better understand coordination structures formed through force and motion signals and potentially develop a robot control model or algorithm to improve physical human-machine coordination.

2018 ASU Robotics Symposium

Physical Human-Machine Coordination Research Paper

References

Coco, M. I., & Dale, R. (2014). Cross-recurrence quantification analysis of categorical and continuous time series: An R package. Frontiers in Psychology, 5(JUN), 1–14. https://doi.org/10.3389/fpsyg.2014.00510

Finch, H., Wijnants, M., Giuliani, A., Wallot, S., & Leonardi, G. (2018). Analyzing Multivariate Dynamics Using Cross-Recurrence Quantification Analysis (CRQA), Diagonal-Cross-Recurrence Profiles (DCRP), and Multidimensional Recurrence Quantification Analysis (MdRQA)-A Tutorial in R. Analyzing Multivariate Dynamics Using Cross-. https://doi.org/10.3389/fpsyg.2018.02232

Jarrassé, N., Sanguineti, V., & Burdet, E. (2014). Slaves no longer: Review on role assignment for human-robot joint motor action. Adaptive Behavior, 22(1), 70–82. https://doi.org/10.1177/1059712313481044

Jarrassé, N., Charalambous, T., and Burdet, E., 2012, “A Framework to Describe, Analyze and Generate Interactive Motor Behaviors,” PLoS One, 7(11), p. e49945

Knoblich, G., Butterfill, S., & Sebanz, N. (2011). Psychological Research on Joint Action. Theory and Data. Psychology of Learning and Motivation - Advances in Research and Theory (Vol. 54). https://doi.org/10.1016/B978-0-12-385527-5.00003-6

Lee, J. D., and See, K. A., 2004, “Trust in Automation: Designing for Appropriate Reliance,” Hum. Factors, 46(1), pp. 50–80.

Mörtl, A., Lawitzky, M., Kucukyilmaz, A., Sezgin, M., Basdogan, C., & Hirche, S. (2012). The role of roles: Physical cooperation between humans and robots. International Journal of Robotics Research, 31(13), 1656–1674. https://doi.org/10.1177/0278364912455366

Reed, K. B. (2005). Haptic cooperation between people, and between people and machines. Caderno CRH, 21(53), 239–252. https://doi.org/10.1109/IROS.2006.282489

K. B. Reed, J. Patton, and M. Peshkin, “Replicating human-human physical interaction,” in Proceedings of the 2007 IEEE International Conference on Robotics and Automation. IEEE, 2007, pp. 3615–3620.

Sadrfaridpour, B., Saeidi, H., Burke, J., Madathil, K., and Wang, Y., 2016, “Modeling and Control of Trust in Human-Robot Collaborative Manufacturing,” Robust Intelligence and Trust in Autonomous Systems, Springer, Boston, MA, pp. 115–114.

Sebanz, N., Bekkering, H., & Knoblich, G. (2006). Joint action: Bodies and minds moving together. Trends in Cognitive Sciences, 10(2), 70–76. https://doi.org/10.1016/j.tics.2005.12.009

Van der Wel, R. P. R. D., Knoblich, G., & Sebanz, N. (2011). Let the Force Be With Us: Dyads Exploit Haptic Coupling for Coordination. Journal of Experimental Psychology: Human Perception and Performance. https://doi.org/10.1037/a0022337

Wallot, S. (2018). Multidimensional Cross-Recurrence Quantification Analysis (MdCRQA)-A Method for Quantifying Correlation between Multivariate Time-Series: Multidimensional Cross-Recurrence Quantification Analysis (MdCRQA)-A Method for Quantifying Correlation between Multi. https://doi.org/10.1080/00273171.2018.1512846

Webber, C.L., & Zbilut, J.P. (2004). 2 Recurrence Quantification Analysis of Nonlinear Dynamical Systems.