Predictive Modeling of Periodic Behavior for Human-Robot Symbiotic Walking

Clark, Geoffrey A.; Campbell, Joseph; Sorkhabadi, Seyed Mostafa Rezayat; Zhang, Wenlong; Amor, Heni Ben

doi:10.1109/icra40945.2020.9196676

Cited by 7 publications

(6 citation statements)

References 22 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Human Collaborator [98][99][100][101][102] Multi-Platform [101,[103][104][105] Infrastructural Sensors [106][107][108] Asset Integrity Inspection [109][110][111] System of Systems [112][113][114] Figure 13.7: Spectator view from a HoloLens where the user is utilizing the HoloLens to interact with a model of an offshore wind farm and examine the effects on the windows [14].…”

Section: Symbiotic Relationship Referencesmentioning

confidence: 99%

The Future Workplace: A Symbiotic System of Systems Environment

McConnell¹,

Mitchell²,

Donaldson³

et al. 2021

Cyber-Physical Systems

View full text Add to dashboard Cite

show abstract

Section: Symbiotic Relationship Referencesmentioning

confidence: 99%

The Future Workplace: A Symbiotic System of Systems Environment

McConnell¹,

Mitchell²,

Donaldson³

et al. 2021

Cyber-Physical Systems

View full text Add to dashboard Cite

show abstract

“…Previous applications of IPs for human-robot interaction include, for example, collaborative assembly [25], throwing and catching [7], or electromyography based object hand over [26], show that IPs can infer modeled parameters including robot joint angles given a set of observed partial trajectories. A first attempt at using IPs for robotic prostheses is presented in [8], where control signals were generated in a purely reactive fashion by iteratively conditioning on the sensor readings. The approach is particularly powerful when dealing with noisy sensing and performance gracefully degrades when one or more sensors unexpectedly break down; however it does not take into account the future ramifications of a selected action.…”

Section: Related Workmentioning

confidence: 99%

“…In this paper, we build upon previous work on IP, but make critical extensions to allow for planning and biomechanically-safe, predictive control. Going beyond reactive control [25,8,7], we show how model-predictive control (MPC) [27,28] can be implemented with IPs as the underlying data-driven model. The resulting approach has powerful inference capabilities that allow for predictive biomechanics: the prediction and control of future, unobserved biomechanical states of the human-robot system from current sensor readings.…”

Section: Related Workmentioning

confidence: 99%

“…In the first mode, we only use the inference mechanisms of IPs to generate reactive control values. This can be seen as a behavioral cloning approach and corresponds to the method introduced in [8]. In addition we evaluated our MPIP variant with a cost function that minmizes a.)…”

Section: Experiments 1: Simulationmentioning

confidence: 99%

“…To this end, we discuss a unified approach to (a) human state estimation and prediction, (b) inference of unobserved biomechanical variables, and (c) model-predictive control in support of the above ergonomic objectives and constraints. Building upon existing literature and prior work on probabilistic human motion modelling [7,8], we present the application of interaction primitives for learning predictive models of biomechanics and control in a lower-limb prosthesis. An important feature of this method is the ability to reduce the cooperative system (human and robot) to a latent space using basis function decomposition, in which temporal features are probabilistically related to one another through their covariance.…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Learning Predictive Models for Ergonomic Control of Prosthetic Devices

Clark¹,

Campbell²,

Amor³

2020

Preprint

Self Cite

View full text Add to dashboard Cite

We present Model-Predictive Interaction Primitives -a robot learning framework for assistive motion in human-machine collaboration tasks which explicitly accounts for biomechanical impact on the human musculoskeletal system. First, we extend Interaction Primitives to enable predictive biomechanics: the prediction of future biomechanical states of a human partner conditioned on current observations and intended robot control signals. In turn, we leverage this capability within a model-predictive control strategy to identify the future ergonomic and biomechanical ramifications of potential robot actions. Optimal control trajectories are selected so as to minimize future physical impact on the human musculoskeletal system. We empirically demonstrate that our approach minimizes knee or muscle forces via generated control actions selected according to biomechanical cost functions. Experiments are performed in synthetic and real-world experiments involving powered prosthetic devices.

show abstract

Symbiotic System of Systems Design for Safe and Resilient Autonomous Robotics in Offshore Wind Farms

Mitchell¹,

Blanche

Zaki

et al. 2021

IEEE Access

View full text Add to dashboard Cite

To reduce Operation and Maintenance (O&M) expenditure on offshore wind farms, wherein 80% of the cost relates to deploying personnel, the offshore wind sector looks to advances in Robotics and Artificial Intelligence (RAI) for solutions. Barriers to residential Beyond Visual Line of Sight (BVLOS) autonomy as a service, include operational challenges in run-time safety compliance, reliability and resilience, due to the complexities of dealing with known and unknown risk in dynamic environments. In this paper we incorporate a Symbiotic System Of Systems Approach (SSOSA) that uses a Symbiotic Digital Architecture (SDA) to provide a cyber physical orchestration of enabling technologies. Implementing a SSOSA enables Cooperation, Collaboration and Corroboration (C 3 ), as to address run-time verification of safety, reliability and resilience during autonomous missions. Our SDA provides a means to synchronize distributed digital models of the robot, environment and infrastructure. Through the coordinated bidirectional communication network of the SDA, the remote human operator has improved visibility and understanding of the mission profile. We evaluate our SSOSA in an asset inspection mission within a confined operating environment. Demonstrating the ability of our SSOSA to overcome safety, reliability and resilience challenges. The SDA supports lifecycle learning and co-evolution with knowledge sharing across the interconnected systems. Our results evaluate both sudden and gradual faults, as well as unknown events, that may jeopardize an autonomous mission. Using distributed and coordinated decision making, SSOSA enhances the analysis of the mission status, which includes diagnostics of critical sub-systems within the resident robot. This evaluation demonstrates that the SSOSA provides enhanced run-time operational resilience and safety compliance to BVLOS autonomous missions. SSOSA has the potential to be a highly transferable methodology to other mission scenarios and technologies, providing a pathway to implementing scalable autonomy as a service.

show abstract

Predictive Modeling of Periodic Behavior for Human-Robot Symbiotic Walking

Cited by 7 publications

References 22 publications

The Future Workplace: A Symbiotic System of Systems Environment

The Future Workplace: A Symbiotic System of Systems Environment

Learning Predictive Models for Ergonomic Control of Prosthetic Devices

Symbiotic System of Systems Design for Safe and Resilient Autonomous Robotics in Offshore Wind Farms

Contact Info

Product

Resources

About