Human-in-the-Loop Optimization of Shared Autonomy in Assistive Robotics

By Deepak Edakkattil Gopinath, Siddarth Jain, Brenna D. Argall. 2017

177 citations recorded by the source corpus.

In this paper, we propose a mathematical framework which formalizes user-driven customization of shared autonomy in assistive robotics as a nonlinear optimization problem. Our insight is to allow the end-user, rather than relying on standard optimization techniques, to perform the optimization procedure, thereby allowing us to leave the exact nature of the cost function indeterminate. We ground our formalism with an interactive optimization procedure that customizes control sharing using an assistive robotic arm. We also present a pilot study that explores interactive optimization with end-users. This study was performed with 17 subjects (4 with spinal cord injury, 13 without injury). Results show all subjects were able to converge to an assistance paradigm, suggesting the existence of optimal solutions. Notably, the amount of assistance was not always optimized for task performance. Instead, some subjects favored retaining more control during the execution over better task performance. The study supports the case for user-driven customization and provides guidance for its continued development and study.

Subfields: Cognitive Neuroscience, Rehabilitation, Reinforcement Learning, Human-Computer Interaction

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Cognitive Neuroscience

In Rehabilitation

In Reinforcement Learning

In Human-Computer Interaction

Citations in the corpus

LILA: Language-Informed Latent Actions (2021, methodology): “Shared autonomy casts robot control as a collaborative process between humans and robots [2, 4, 16, 22].”
Shared autonomy via hindsight optimization for teleoperation and teaming (2017, result): “Blending is one of the most used shared control teleo-pration paradigms due to computational eﬃciency, simplicity, and empirical eﬀectiveness (Li et al., 2011; Carlson and Demiris, 2012; Dragan and Srinivasa, 2013b; Muelling et al., 2015; Gopinath et al., 2016).”
A Review of User Interface Design for Interactive Machine Learning (2018, background): “Beyond human–computer interactions, IML related concepts have also been explored in human–robot interaction research (Losing et al. 2015; Javdani et al. 2016; Gopinath et al. 2017).”
Controlling Assistive Robots with Learned Latent Actions (2019, background): “Because of the physical limitations of their users, today’s assistive arms utilize low-dimensional control interfaces, such as joysticks [2].”
A Survey of Human-in-the-loop for Machine Learning (2021, methodology): “[187] X 2016 X X Shared autonomy Holzinger [188] X 2016 X X X X X Interactive system application Conor Walsh [189] X 2018 X X Soft wearable robots Sutton et al.”
Probabilistic Human Intent Recognition for Shared Autonomy in Assistive Robotics (2019, methodology): “1 (RBII-1 and RBII-2), for comparative purposes we also implement two approaches to intent inference utilized in previous shared autonomy works—amnesic inference [2, 9, 13] and memory-based inference [8, 9, 20, 30].”
Learning latent actions to control assistive robots (2021, unknown): “Similar to [16,25,20], we assume access to a discrete set of high-level goals G that the human may want to reach.”
Autonomy in Rehabilitation Robotics: An Intersection (2018, unknown): “This uniqueness strongly suggests a need for customized, even individualized, assistance solutions—and specifically in regards to how control is shared (141).”
Shared Autonomy with Learned Latent Actions (2020, background): “Many shared autonomy algorithms provide robotic assistance to users who must reach for a goal objects in the environment [8, 16, 22, 11, 2, 13].”