Learning a Control Policy for Fall Prevention on an Assistive Walking Device

@article{Kumar2020LearningAC,
  title={Learning a Control Policy for Fall Prevention on an Assistive Walking Device},
  author={Visak C. V. Kumar and Sehoon Ha and Gergory Sawicki and C. Karen Liu},
  journal={2020 IEEE International Conference on Robotics and Automation (ICRA)},
  year={2020},
  pages={4833-4840}
}
Fall prevention is one of the most important components in senior care. We present a technique to augment an assistive walking device with the ability to prevent falls. Given an existing walking device, our method develops a fall predictor and a recovery policy by utilizing the onboard sensors and actuators. The key component of our method is a robust human walking policy that models realistic human gait under a moderate level of perturbations. We use this human walking policy to provide… 

Figures from this paper

Robust Walking Control of a Lower Limb Rehabilitation Exoskeleton Coupled with a Musculoskeletal Model via Deep Reinforcement Learning
Background: Few studies have systematically investigated robust controllers for lower limb rehabilitation exoskeletons (LLREs) that can safely and effectively assist users with a variety of
A model-free deep reinforcement learning approach for control of exoskeleton gait patterns
TLDR
This control technique is the first to accurately track any gait pattern desired in physiotherapy without requiring a predefined dynamics model and is robust to varying post-stroke individuals’ baseline gait patterns and their interactions and perturbations.
Reinforcement Learning and Control of a Lower Extremity Exoskeleton for Squat Assistance
TLDR
This work proposes a new, reinforcement learning-based, motion controller for a lower extremity rehabilitation exoskeleton, aiming to perform collaborative squatting exercises with efficiency, stability, and strong robustness.
End-to-End Deep Reinforcement Learning for Exoskeleton Control
TLDR
This paper presents for the first time an end-to-end model-free deep reinforcement learning method for an exoskeleton that can learn to follow a desired gait pattern, while considering a user’s existing gait patterns and being robust to their perturbations and interactions.
A Lower-Body Exoskeleton Simulation and Control Framework based on Deep Reinforcement Learning by Lowell Rose A thesis submitted in conformity with the requirements
A Lower-Body Exoskeleton Simulation and Control Framework based on Deep Reinforcement Learning Lowell Rose Master of Applied Science Department of Mechanical and Industrial Engineering University of
Preventing falls: the use of machine learning for the prediction of future falls in individuals without history of fall
TLDR
The current advancements and perspectives that may arise from the combination of affordable and easy-to-use non-wearable systems, wearable systems with appropriate machine learning analytics, as well as the efforts to address these challenges are summarized.
Latest Research Trends in Fall Detection and Prevention Using Machine Learning: A Systematic Review
TLDR
This paper presents the latest research trends in fall detection and prevention systems using Machine Learning (ML) algorithms and uses recent studies and analyzes datasets, age groups, ML algorithms, sensors, and location.
Modelling Human Kinetics and Kinematics during Walking using Reinforcement Learning
TLDR
This work develops an automated method to generate 3D human walking motion in simulation which is comparable to real-world human motion and shows that the method generalizes well across human-subjects with different kinematic structure and gait-characteristics.
Error-Aware Policy Learning: Zero-Shot Generalization in Partially Observable Dynamic Environments
TLDR
A novel approach to tackle a sim-to-real problem by developing policies capable of adapting to new environments, in a zero-shot manner, using an error-aware policy (EAP) that is explicitly made aware of the effect of unobservable factors during training.

References

SHOWING 1-10 OF 41 REFERENCES
Path Control: A Method for Patient-Cooperative Robot-Aided Gait Rehabilitation
TLDR
A patient-cooperative strategy that allows patients to influence the timing of their leg movements along a physiologically meaningful path and allow patients to train walking while being helped rather than controlled by the robot.
Toward Balance Recovery With Leg Prostheses Using Neuromuscular Model Control
TLDR
An alternative control policy for powered transfemoral prostheses that considers the global leg function and is based on a neuromuscular model of human locomotion is explored, providing a potential avenue for future development of control policies that help to improve amputee balance recovery.
An ecologically-controlled exoskeleton can improve balance recovery after slippage
TLDR
This study shows a novel ecological approach relying on a wearable robotic device (the Active Pelvis Orthosis, APO) aimed at facilitating balance recovery after unexpected slippages and demonstrates the potential of closed-loop controlled wearable robots to assist elderly and disabled subjects and to improve their quality of life.
Virtual Constraint Control of a Powered Prosthetic Leg: From Simulation to Experiments With Transfemoral Amputees
TLDR
This paper derives exact and approximate control laws for a partial feedback linearization to enforce virtual constraints on a prosthetic leg and encodes a human-inspired invariance property called effective shape into virtual constraints for the stance period.
Hybrid Control of the Berkeley Lower Extremity Exoskeleton (BLEEX)
TLDR
An improved control scheme is presented that adds robustness to changing BLEEX backpack payload that is difficult to obtain and subject to change as payload is added and removed.
Simulating balance recovery responses to trips based on biomechanical principles
TLDR
This paper designs controllers for the trip recovery responses that occur during walking based on the available biomechanical literature and data captured from human subjects in the laboratory.
Stepping in the direction of the fall: the next foot placement can be predicted from current upper body state in steady-state walking
TLDR
It is shown that a linear function of the hip position and velocity state during mid-stance explains over 80% of the next lateral foot position variance, consistent with (but not proving) lateral stabilization using foot placement.
Design and Control of a Powered Hip Exoskeleton for Walking Assistance
TLDR
The research results are presented into a powered hip exoskeleton (PH-EXOS) designed to provide locomotive assistance to individuals with walking impediments and control approaches in reducing walking effort.
Design and Experimental Verification of Hip Exoskeleton With Balance Capacities for Walking Assistance
TLDR
A novel, high-power, self-balancing, and passively and software-controlled actively compliant hip exoskeleton that can assist with movement and maintain balance in both the sagittal and frontal planes is developed.
Design and Control of the MINDWALKER Exoskeleton
TLDR
The design, control, and preliminary evaluation of a novel exoskeleton, MINDWALKER, which has a novel step-width adaptation algorithm to stabilize lateral balance and which tested on both healthy subjects and paraplegics showed that all users could successfully trigger steps by CoM displacement.
...
...