Controllers as filters: Noise-driven swing-up control based on Maxwell's demon

Tzorakoleftherakis, Emmanouil; Murphey, Todd D.

doi:10.1109/cdc.2015.7402901

Cited by 10 publications

(9 citation statements)

References 22 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…An algorithm for filtering control inputs was proposed in Tzorakoleftherakis and Murphey (2015) for noise-driven swing-up problems based on the hypothesis that noisy inputs can be a rich source of control authority if filtered in a meaningful task-specific way. This filter was implemented by combining a controller and a filter into a single computational unit that cancels noise samples not driving the system towards a desired control direction.…”

Section: Prior Workmentioning

confidence: 99%

Task-based hybrid shared control for training through forceful interaction

Fitzsimons

Kalinowska

Dewald

et al. 2020

The International Journal of Robotics Research

View full text Add to dashboard Cite

Despite the fact that robotic platforms can provide both consistent practice and objective assessments of users over the course of their training, there are relatively few instances where physical human–robot interaction has been significantly more effective than unassisted practice or human-mediated training. This article describes a hybrid shared control robot, which enhances task learning through kinesthetic feedback. The assistance assesses user actions using a task-specific evaluation criterion and selectively accepts or rejects them at each time instant. Through two human subject studies (total [Formula: see text]), we show that this hybrid approach of switching between full transparency and full rejection of user inputs leads to increased skill acquisition and short-term retention compared with unassisted practice. Moreover, we show that the shared control paradigm exhibits features previously shown to promote successful training. It avoids user passivity by only rejecting user actions and allowing failure at the task. It improves performance during assistance, providing meaningful task-specific feedback. It is sensitive to initial skill of the user and behaves as an “assist-as-needed” control scheme, adapting its engagement in real time based on the performance and needs of the user. Unlike other successful algorithms, it does not require explicit modulation of the level of impedance or error amplification during training and it is permissive to a range of strategies because of its evaluation criterion. We demonstrate that the proposed hybrid shared control paradigm with a task-based minimal intervention criterion significantly enhances task-specific training.

show abstract

Section: Prior Workmentioning

confidence: 99%

Task-based hybrid shared control for training through forceful interaction

Fitzsimons

Kalinowska

Dewald

et al. 2020

The International Journal of Robotics Research

View full text Add to dashboard Cite

show abstract

“…In this experiment, we implemented a form of assistance that can convert pure noise input into a successful task execution by comparing the noise input to that of an optimal controller [28]. The assistance acts as a filter similar to that described in [28] and [29], such that if user inputs agree with the optimal controller, user input is not modified by the interface. When user inputs do not agree, the robot physically rejects the input, providing feedback but not guidance.…”

Section: Resultsmentioning

confidence: 99%

“…The assistance algorithm used in these experiments was Maxwell’s Demon Algorithm (MDA). The MDA algorithm was proposed in [28] for noise-driven non-linear control based on the hypothesis that noisy inputs can be a rich source of control authority if filtered in a task-specific way. The MDA filter was implemented by combining a controller and a filter into a single computational unit that cancels noise samples not driving the system to the desired control direction.…”

Section: Methodsmentioning

confidence: 99%

Ergodicity reveals assistance and learning from physical human-robot interaction

et al. 2019

Self Cite

View full text Add to dashboard Cite

This paper applies information theoretic principles to the investigation of physical human-robot interaction. Drawing from the study of human perception and neural encoding, information theoretic approaches offer a perspective that enables quantitatively interpreting the body as an information channel, and bodily motion as an information-carrying signal. We show that ergodicity, which can be interpreted as the degree to which a trajectory encodes information about a task, correctly predicts changes due to reduction of a person’s existing deficit or the addition of algorithmic assistance. The measure also captures changes from training with robotic assistance. Other common measures for assessment failed to capture at least one of these effects. This information-based interpretation of motion can be applied broadly, in the evaluation and design of human-machine interactions, in learning by demonstration paradigms, or in human motion analysis.

show abstract

“…In this work, we allocate control using a filter (Tzorakoleftherakis and Murphey, 2015) described more thoroughly in Section 3.3. Our control allocation strategy is similar in practice to virtual fixtures and virtual guides , techniques that are common in the haptics literature (Forsyth and MacLean, 2005; Griffiths and Gillespie, 2005).…”

Section: Background and Related Workmentioning

confidence: 99%

“…We use a geometric signal filter that is capable of dynamically shifting which partner is in control at any given instant based on optimality criteria. This technique is known as Maxwell’s Demon Algorithm (MDA) (Tzorakoleftherakis and Murphey, 2015). Our specific implementation of MDA is detailed in Algorithm 1 where u h is the control input from the human operator, u a is the control produced by the autonomy, and u is applied to the dynamic system.…”

Section: Mbscmentioning

confidence: 99%

Data-driven Koopman operators for model-based shared control of human–machine systems

Broad

Abraham

Murphey

et al. 2020

The International Journal of Robotics Research

View full text Add to dashboard Cite

We present a data-driven shared control algorithm that can be used to improve a human operator’s control of complex dynamic machines and achieve tasks that would otherwise be challenging, or impossible, for the user on their own. Our method assumes no a priori knowledge of the system dynamics. Instead, both the dynamics and information about the user’s interaction are learned from observation through the use of a Koopman operator. Using the learned model, we define an optimization problem to compute the autonomous partner’s control policy. Finally, we dynamically allocate control authority to each partner based on a comparison of the user input and the autonomously generated control. We refer to this idea as model-based shared control (MbSC). We evaluate the efficacy of our approach with two human subjects studies consisting of 32 total participants (16 subjects in each study). The first study imposes a linear constraint on the modeling and autonomous policy generation algorithms. The second study explores the more general, nonlinear variant. Overall, we find that MbSC significantly improves task and control metrics when compared with a natural learning, or user only, control paradigm. Our experiments suggest that models learned via the Koopman operator generalize across users, indicating that it is not necessary to collect data from each individual user before providing assistance with MbSC. We also demonstrate the data efficiency of MbSC and, consequently, its usefulness in online learning paradigms. Finally, we find that the nonlinear variant has a greater impact on a user’s ability to successfully achieve a defined task than the linear variant.

show abstract

Controllers as filters: Noise-driven swing-up control based on Maxwell's demon

Cited by 10 publications

References 22 publications

Task-based hybrid shared control for training through forceful interaction

Task-based hybrid shared control for training through forceful interaction

Ergodicity reveals assistance and learning from physical human-robot interaction

Data-driven Koopman operators for model-based shared control of human–machine systems

Contact Info

Product

Resources

About