Review of machine learning methods in soft robotics

Kim, Daekyum; Kim, Sanghun; Kim, Taekyoung; Kang, Brian Byunghyun; Lee, Minhyuk; Park, Wookeun; Ku, Subyeong; Kim, DongWook; Kwon, Junghan; Lee, Ho-Chang; Bae, Joonbum; Park, Yong‐Lae; Cho, Kyu-Jin; Jo, Sungho

doi:10.1371/journal.pone.0246102

Cited by 138 publications

(93 citation statements)

References 131 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…This article will review the recent trends in design, sensing and control of wearable soft robots, which supplements and extends several recent review papers on textile-based wearable robots [31], rehabilitation robots powered by pneumatic muscles [32], machine learning methods in soft robotics [33], and control strategies for soft continuum robotic manipulators [34]. This paper attempts to present a holistic view of different aspects required to integrate wearable soft robots with human users, which include actuation mechanisms to drive wearable soft robots, designs and algorithms of soft sensors, controller syntheses for wearable soft robots to autonomously collaborate with human users, and human evaluation results.…”

Section: Introductionmentioning

confidence: 90%

Human Assistance and Augmentation with Wearable Soft Robotics: a Literature Review and Perspectives

et al. 2021

View full text Add to dashboard Cite

Purpose of ReviewWearable soft robotics has demonstrated many unique capabilities in human assistance and augmentation. This paper discusses the recent trends and remaining challenges in designing soft actuators, soft sensors, and controllers for wearable soft robots.Recent Findings Different actuation mechanisms for wearable systems were investigated in this review. We include a synopsis on the design of soft sensors and algorithms for sensor fusion and discuss the recent trends for sensing simultaneous deformations and implementing machine learning techniques for soft sensor fusion. We also present a discussion on layered controller design and the evaluation metrics for experiments with healthy and impaired subjects. SummaryWe review three commonly used soft actuation mechanisms and provide their characteristics, in which the balance between material stiffness and functionality is still a challenge. We present the advances in soft sensor design and discuss the challenges in fusion algorithms for wearable soft sensors, such as inter-subject adaptability and sensor location movement. We also discuss the model-based and model-free controller design together with the evaluation criteria for healthy and impaired users, in which autonomous operation for personalized assistance is still an open challenge for controller design.

show abstract

Section: Introductionmentioning

confidence: 90%

Human Assistance and Augmentation with Wearable Soft Robotics: a Literature Review and Perspectives

et al. 2021

View full text Add to dashboard Cite

show abstract

“…The model-free methods for modeling soft robots mainly use data-driven machine learning and deep learning techniques to find a mapping between the inputs and outputs of the soft system [108]. Input actuation signals and robot states can be obtained by sensors, either embedded or external visual tracking sensors.…”

Section: Modelingmentioning

confidence: 99%

Underwater Soft Robotics: A Review of Bioinspiration in Design, Actuation, Modeling, and Control

et al. 2022

View full text Add to dashboard Cite

Nature and biological creatures are some of the main sources of inspiration for humans. Engineers have aspired to emulate these natural systems. As rigid systems become increasingly limited in their capabilities to perform complex tasks and adapt to their environment like living creatures, the need for soft systems has become more prominent due to the similar complex, compliant, and flexible characteristics they share with intelligent natural systems. This review provides an overview of the recent developments in the soft robotics field, with a focus on the underwater application frontier.

show abstract

“…Thus, to discriminate different sensory feedback, one could envision innovative machine learning approaches relying on different domains (e.g., time and frequency). [389] In conclusion, multifunctional elements ("material intelligence"), mechanical properties for various geometrical conformations ("mechanical intelligence"), and fast learning and responsive processes ("learning intelligence"), combined with ecofriendly characteristics, [72,390] are the basement to endow the future soft robotic platforms with an embodied intelligence able to adapt and coexist with ecological environments and human users (Figure 12). Goffredo Giordano received his M.Sc.…”

Section: Perspective and Conclusionmentioning

confidence: 99%

A Perspective on Cephalopods Mimicry and Bioinspired Technologies toward Proprioceptive Autonomous Soft Robots

Giordano

Carlotti

Mazzolai

2021

Adv Materials Technologies

View full text Add to dashboard Cite

scientific community. [1][2][3][4][5][6][7] Regardless, every time a living being reveals novel adaptive and dynamically reactive mimicry behavior, it inspires and fosters futuristic and unexpected technological outcomes. [8][9][10][11][12] At the biological level, the visual crypsis is the ability of a species to resemble the surroundings by matching the coloration and the geometrical patterns of the habitat. In this sense, a living being can control its appearance optically (thanks to arranged and optimized structures at the mesoscopic scale, by means of pigmentation, or emissive elements) and morphologically (it can show wrinkles and textures on the body to evade detection or observation). [13][14][15][16][17][18] Both these mechanisms are characterized by a time response that ranges from milliseconds to hundreds of seconds. In nature, several species take advantage of cryptic abilities, as, e.g., in cephalopods, [7] crustaceans, [19] reptilians, [1,20,21] insects, [22,23] birds, [24,25] shells, [26,27] plants, [28,29] among many. The biotic color changing and body patterning relate to reproductive, communicative, and defensive and/or predatory strategies. Unfortunately, the nervous or central control-chain system that steers these behaviors in animals and plants, is still somehow fogged for scientists. [7,[30][31][32] A complete knowledge about their central information process systems, can open to astonishing developments in many scientific branches, from neurobiology [33,34] to quantum biology. [35] Undoubtedly, the most debated case of study in the natural world are cephalopods, which not only are highly evolved and specialized in fast adaptive color changing displays, but also are able to actuate their skin to generate 3D patterns, when exposed to specific mechanical, thermal, optical, or chemical stimuli. Soft muscular arrangements, [36][37][38] spatially distributed and expandable absorbing components (namely chromatophores), [39,40] iridescent elements (namely irido phores), [41,42] and bright white scatterers (namely leucophores) [43] are responsible for textural, postural and color changing. Moreover, octopuses are a remarkable intelligent species, able, for example, to open a jar or avoid predators in the order of seconds. [44] Thus, cephalopods are usually regarded as a perfect example of embodied intelligence, [45] thanks to the symbiosis between their body's mechanics and morphology, and the distributed sensory neuro-motor-control system. Their "learning", "mechanical" and "material intelligence" will be our lodestars along this review, resulting in a fascinating connection between Octopus skin is an amazing source of inspiration for bioinspired sensors, actuators and control solutions in soft robotics. Soft organic materials, biomacromolecules and protein ingredients in octopus skin combined with a distributed intelligence, result in adaptive displays that can control emerging optical behavior, and 3D surface textures with rough geometries, with a remarkably high control speed (≈ms). To be able ...

show abstract

Review of machine learning methods in soft robotics

Cited by 138 publications

References 131 publications

Human Assistance and Augmentation with Wearable Soft Robotics: a Literature Review and Perspectives

Human Assistance and Augmentation with Wearable Soft Robotics: a Literature Review and Perspectives

Underwater Soft Robotics: A Review of Bioinspiration in Design, Actuation, Modeling, and Control

A Perspective on Cephalopods Mimicry and Bioinspired Technologies toward Proprioceptive Autonomous Soft Robots

Contact Info

Product

Resources

About