1
|
Wang X, Santos VJ. Gaze-Based Shared Autonomy Framework With Real-Time Action Primitive Recognition for Robot Manipulators. IEEE Trans Neural Syst Rehabil Eng 2023; 31:4306-4317. [PMID: 37906485 DOI: 10.1109/tnsre.2023.3328888] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [MESH Headings] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Indexed: 11/02/2023]
Abstract
Robots capable of robust, real-time recognition of human intent during manipulation tasks could be used to enhance human-robot collaboration for activities of daily living. Eye gaze-based control interfaces offer a non-invasive way to infer intent and reduce the cognitive burden on operators of complex robots. Eye gaze is traditionally used for "gaze triggering" (GT) in which staring at an object, or sequence of objects, triggers pre-programmed robotic movements. We propose an alternative approach: a neural network-based "action prediction" (AP) mode that extracts gaze-related features to recognize, and often predict, an operator's intended action primitives. We integrated the AP mode into a shared autonomy framework capable of 3D gaze reconstruction, real-time intent inference, object localization, obstacle avoidance, and dynamic trajectory planning. Using this framework, we conducted a user study to directly compare the performance of the GT and AP modes using traditional subjective performance metrics, such as Likert scales, as well as novel objective performance metrics, such as the delay of recognition. Statistical analyses suggested that the AP mode resulted in more seamless robotic movement than the state-of-the-art GT mode, and that participants generally preferred the AP mode.
Collapse
|
2
|
Taylor S, Wang M, Jeon M. Reliable and transparent in-vehicle agents lead to higher behavioral trust in conditionally automated driving systems. Front Psychol 2023; 14:1121622. [PMID: 37275735 PMCID: PMC10232983 DOI: 10.3389/fpsyg.2023.1121622] [Citation(s) in RCA: 1] [Impact Index Per Article: 1.0] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Figures] [Journal Information] [Subscribe] [Scholar Register] [Received: 12/12/2022] [Accepted: 05/02/2023] [Indexed: 06/07/2023] Open
Abstract
Trust is critical for human-automation collaboration, especially under safety-critical tasks such as driving. Providing explainable information on how the automation system reaches decisions and predictions can improve system transparency, which is believed to further facilitate driver trust and user evaluation of the automated vehicles. However, what the optimal level of transparency is and how the system communicates it to calibrate drivers' trust and improve their driving performance remain uncertain. Such uncertainty becomes even more unpredictable given that the system reliability remains dynamic due to current technological limitations. To address this issue in conditionally automated vehicles, a total of 30 participants were recruited in a driving simulator study and assigned to either a low or a high system reliability condition. They experienced two driving scenarios accompanied by two types of in-vehicle agents delivering information with different transparency types: "what"-then-wait (on-demand) and "what + why" (proactive). The on-demand agent provided some information about the upcoming event and delivered more information if prompted by the driver, whereas the proactive agent provided all information at once. Results indicated that the on-demand agent was more habitable, or naturalistic, to drivers and was perceived with faster system response speed compared to the proactive agent. Drivers under the high-reliability condition complied with the takeover request (TOR) more (if the agent was on-demand) and had shorter takeover times (in both agent conditions) compared to those under the low-reliability condition. These findings inspire how the automation system can deliver information to improve system transparency while adapting to system reliability and user evaluation, which further contributes to driver trust calibration and performance correction in future automated vehicles.
Collapse
Affiliation(s)
- Skye Taylor
- Mind Music Machine Lab, Grado Department of Industrial and Systems Engineering, Virginia Tech, Blacksburg, VA, United States
- Link Lab, Department of Systems and Information Engineering, University of Virginia, Charlottesville, VA, United States
| | - Manhua Wang
- Mind Music Machine Lab, Grado Department of Industrial and Systems Engineering, Virginia Tech, Blacksburg, VA, United States
| | - Myounghoon Jeon
- Mind Music Machine Lab, Grado Department of Industrial and Systems Engineering, Virginia Tech, Blacksburg, VA, United States
| |
Collapse
|
3
|
Kraus J, Babel F, Hock P, Hauber K, Baumann M. The trustworthy and acceptable HRI checklist (TA-HRI): questions and design recommendations to support a trust-worthy and acceptable design of human-robot interaction. GIO-GRUPPE-INTERAKTION-ORGANISATION-ZEITSCHRIFT FUER ANGEWANDTE ORGANISATIONSPSYCHOLOGIE 2022. [DOI: 10.1007/s11612-022-00643-8] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Indexed: 10/16/2022]
Abstract
AbstractThis contribution to the journal Gruppe. Interaktion. Organisation. (GIO) presents a checklist of questions and design recommendations for designing acceptable and trustworthy human-robot interaction (HRI). In order to extend the application scope of robots towards more complex contexts in the public domain and in private households, robots have to fulfill requirements regarding social interaction between humans and robots in addition to safety and efficiency. In particular, this results in recommendations for the design of the appearance, behavior, and interaction strategies of robots that can contribute to acceptance and appropriate trust. The presented checklist was derived from existing guidelines of associated fields of application, the current state of research on HRI, and the results of the BMBF-funded project RobotKoop. The trustworthy and acceptable HRI checklist (TA-HRI) contains 60 design topics with questions and design recommendations for the development and design of acceptable and trustworthy robots. The TA-HRI Checklist provides a basis for discussion of the design of service robots for use in public and private environments and will be continuously refined based on feedback from the community.
Collapse
|
4
|
Sadeghian H, Naceri A, Haddadin S. Munich Institute of Robotics and Machine Intelligence (MIRMI) Technical University of Munich. Laryngorhinootologie 2022; 101:S186-S193. [PMID: 35605619 DOI: 10.1055/a-1663-0803] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Indexed: 10/18/2022]
Abstract
The application of robotic and intelligent technologies in healthcare is dramatically increasing. The next generation of lightweight and tactile robots have provided a great opportunity to be used for a wide range of applications from medical examination, diagnosis, therapeutic procedures to rehabilitation and assistive robotics. They can potentially outperform current medical procedures by exploiting the com- plementary strengths of humans and computer-based technologies. In this study, the importance of human- robot interaction is discussed and technological re- quirements and challenges in making human-centered robot platforms for medical applications is addressed.
Collapse
Affiliation(s)
- Hamid Sadeghian
- Munich Institute of Robotics and Machine Intelligence (MIRMI), Technische Universität München
| | - Abdeldjallil Naceri
- Munich Institute of Robotics and Machine Intelligence (MIRMI), Technische Universität München
| | - Sami Haddadin
- Munich Institute of Robotics and Machine Intelligence (MIRMI), Technische Universität München
| |
Collapse
|
5
|
Williams J, Fiore SM, Jentsch F. Supporting Artificial Social Intelligence With Theory of Mind. Front Artif Intell 2022; 5:750763. [PMID: 35295867 PMCID: PMC8919046 DOI: 10.3389/frai.2022.750763] [Citation(s) in RCA: 2] [Impact Index Per Article: 1.0] [Reference Citation Analysis] [Abstract] [Grants] [Track Full Text] [Download PDF] [Journal Information] [Subscribe] [Scholar Register] [Received: 07/31/2021] [Accepted: 01/24/2022] [Indexed: 11/13/2022] Open
Abstract
In this paper, we discuss the development of artificial theory of mind as foundational to an agent's ability to collaborate with human team members. Agents imbued with artificial social intelligence will require various capabilities to gather the social data needed to inform an artificial theory of mind of their human counterparts. We draw from social signals theorizing and discuss a framework to guide consideration of core features of artificial social intelligence. We discuss how human social intelligence, and the development of theory of mind, can contribute to the development of artificial social intelligence by forming a foundation on which to help agents model, interpret and predict the behaviors and mental states of humans to support human-agent interaction. Artificial social intelligence will need the processing capabilities to perceive, interpret, and generate combinations of social cues to operate within a human-agent team. Artificial Theory of Mind affords a structure by which a socially intelligent agent could be imbued with the ability to model their human counterparts and engage in effective human-agent interaction. Further, modeling Artificial Theory of Mind can be used by an ASI to support transparent communication with humans, improving trust in agents, so that they may better predict future system behavior based on their understanding of and support trust in artificial socially intelligent agents.
Collapse
Affiliation(s)
- Jessica Williams
- Team Performance Laboratory, University of Central Florida, Institute for Simulation and Training, Orlando, FL, United States
- *Correspondence: Jessica Williams ;
| | - Stephen M. Fiore
- Cognitive Sciences Laboratory, University of Central Florida, Institute for Simulation and Training, Orlando, FL, United States
| | - Florian Jentsch
- Team Performance Laboratory, University of Central Florida, Institute for Simulation and Training, Orlando, FL, United States
| |
Collapse
|
6
|
Future Directions for Human-Centered Transparent Systems for Engine Room Monitoring in Shore Control Centers. JOURNAL OF MARINE SCIENCE AND ENGINEERING 2021. [DOI: 10.3390/jmse10010022] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Track Full Text] [Subscribe] [Scholar Register] [Indexed: 11/17/2022]
Abstract
Many autonomous ship projects have reflected the increasing interest in incorporating the concept of autonomy into the maritime transportation sector. However, autonomy is not a silver bullet, as exemplified by many incidents in the past involving human and machine interaction; rather it introduces new Human Factor (HF) challenges. These challenges are especially critical for Engine Room Monitoring (ERM) in Shore Control Centre (SCCs) due to the system’s complexity and the absence of human senses in the decision-making process. A transparent system is one of the potential solutions, providing a rationale behind its suggestion. However, diverse implementations of transparency schemes have resulted in prevalent inconsistencies in its effects. This literature review paper investigates 17 transparency studies published over the last eight years to identify (a) different approaches to developing transparent systems, (b) the effects of transparency on key HFs, and (c) the effects of information presentation methods and uncertainty information. The findings suggest that the explicit presentation of information could strengthen the benefits of the transparent system and could be promising for performance improvements in ERM tasks in the SCC.
Collapse
|
7
|
Selvaggio M, Cognetti M, Nikolaidis S, Ivaldi S, Siciliano B. Autonomy in Physical Human-Robot Interaction: A Brief Survey. IEEE Robot Autom Lett 2021. [DOI: 10.1109/lra.2021.3100603] [Citation(s) in RCA: 9] [Impact Index Per Article: 3.0] [Reference Citation Analysis] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Indexed: 11/07/2022]
|
8
|
Cini F, Banfi T, Ciuti G, Craighero L, Controzzi M. The relevance of signal timing in human-robot collaborative manipulation. Sci Robot 2021; 6:eabg1308. [PMID: 34550718 DOI: 10.1126/scirobotics.abg1308] [Citation(s) in RCA: 3] [Impact Index Per Article: 1.0] [Reference Citation Analysis] [Abstract] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Indexed: 11/02/2022]
Abstract
To achieve a seamless human-robot collaboration, it is crucial that robots express their intentions without perturbating or interrupting the task that a human partner is performing at that moment. Although it has not received much attention so far, this issue is important when robots assist humans in physical and manipulation tasks. The main question addressed here is whether there is a more appropriate time to inform a human partner that a robot is requesting to pass them an object. This question is posed in a reference scenario where human individuals are involved in a continuous pick-and-place task that cannot be interrupted. Our findings showed that providing a cue at the beginning of a reach-to-grasp movement could severely interfere with the ongoing human action, increasing the number of errors made by humans, slowing down and degrading the smoothness of their arm movement, and deflecting their gaze. These disruptive interferences strongly decreased, until they disappeared, when the robot provided the cue to the human partners shortly after the participants picked up an object, identifying this as the best signaling timing. The results of this work showed how the signaling timing may have a decisive influence on the performances of the human-robot teamwork and contribute to understanding the mechanisms underpinning the phenomenon of cognitive-motor interference in humans.
Collapse
Affiliation(s)
- F Cini
- BioRobotics Institute, Scuola Superiore Sant'Anna, Pisa, Italy.,Scuola Superiore Sant'Anna, Department of Excellence in Robotics & AI, Pisa, Italy
| | - T Banfi
- BioRobotics Institute, Scuola Superiore Sant'Anna, Pisa, Italy.,Scuola Superiore Sant'Anna, Department of Excellence in Robotics & AI, Pisa, Italy
| | - G Ciuti
- BioRobotics Institute, Scuola Superiore Sant'Anna, Pisa, Italy.,Scuola Superiore Sant'Anna, Department of Excellence in Robotics & AI, Pisa, Italy
| | - L Craighero
- University of Ferrara, Department of Neuroscience and Rehabilitation, Ferrara, Italy
| | - M Controzzi
- BioRobotics Institute, Scuola Superiore Sant'Anna, Pisa, Italy.,Scuola Superiore Sant'Anna, Department of Excellence in Robotics & AI, Pisa, Italy
| |
Collapse
|
9
|
Winfield AFT, Booth S, Dennis LA, Egawa T, Hastie H, Jacobs N, Muttram RI, Olszewska JI, Rajabiyazdi F, Theodorou A, Underwood MA, Wortham RH, Watson E. IEEE P7001: A Proposed Standard on Transparency. Front Robot AI 2021; 8:665729. [PMID: 34381820 PMCID: PMC8351056 DOI: 10.3389/frobt.2021.665729] [Citation(s) in RCA: 11] [Impact Index Per Article: 3.7] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Download PDF] [Journal Information] [Subscribe] [Scholar Register] [Received: 02/08/2021] [Accepted: 07/05/2021] [Indexed: 11/13/2022] Open
Abstract
This paper describes IEEE P7001, a new draft standard on transparency of autonomous systems. In the paper, we outline the development and structure of the draft standard. We present the rationale for transparency as a measurable, testable property. We outline five stakeholder groups: users, the general public and bystanders, safety certification agencies, incident/accident investigators and lawyers/expert witnesses, and explain the thinking behind the normative definitions of "levels" of transparency for each stakeholder group in P7001. The paper illustrates the application of P7001 through worked examples of both specification and assessment of fictional autonomous systems.
Collapse
Affiliation(s)
| | - Serena Booth
- Computer Science and AI Laboratory (CSAIL), MIT, Cambridge, MA, United States
| | - Louise A Dennis
- Department of Computer Science, University of Manchester, Manchester, United Kingdom
| | | | - Helen Hastie
- Department of Computer Science, Heriot-Watt University, Edinburgh, United Kingdom
| | - Naomi Jacobs
- ImaginationLancaster, Lancaster Institute for Contemporary Arts, University of Lancaster, Lancaster, United Kingdom
| | | | - Joanna I Olszewska
- School of Computing and Engineering, University of the West of Scotland, Paisley, United Kingdom
| | - Fahimeh Rajabiyazdi
- Department of Mechanical and Industrial Engineering, University of Toronto, Toronto, ON, Canada
| | | | | | - Robert H Wortham
- Department of Electronic and Electrical Engineering, University of Bath, Bath, United Kingdom
| | | |
Collapse
|
10
|
Wallkötter S, Tulli S, Castellano G, Paiva A, Chetouani M. Explainable Embodied Agents Through Social Cues. ACM TRANSACTIONS ON HUMAN-ROBOT INTERACTION 2021. [DOI: 10.1145/3457188] [Citation(s) in RCA: 5] [Impact Index Per Article: 1.7] [Reference Citation Analysis] [Abstract] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Indexed: 10/20/2022]
Abstract
The issue of how to make embodied agents explainable has experienced a surge of interest over the past 3 years, and there are many terms that refer to this concept, such as transparency and legibility. One reason for this high variance in terminology is the unique array of social cues that embodied agents can access in contrast to that accessed by non-embodied agents. Another reason is that different authors use these terms in different ways. Hence, we review the existing literature on explainability and organize it by (1) providing an overview of existing definitions, (2) showing how explainability is implemented and how it exploits different social cues, and (3) showing how the impact of explainability is measured. Additionally, we present a list of open questions and challenges that highlight areas that require further investigation by the community. This provides the interested reader with an overview of the current state of the art.
Collapse
Affiliation(s)
| | - Silvia Tulli
- Uppsala University and INESC-ID-Instituto Superior Técnico, Lisbon, Portugal
| | | | - Ana Paiva
- INESC-ID and Instituto Superior Técnico
| | - Mohamed Chetouani
- Institute for Intelligent Systems and Robotics, CNRS UMR 7222, Sorbonne Université, Paris, France
| |
Collapse
|
11
|
Lillo PD, Arrichiello F, Vito DD, Antonelli G. BCI-Controlled Assistive Manipulator: Developed Architecture and Experimental Results. IEEE Trans Cogn Dev Syst 2021. [DOI: 10.1109/tcds.2020.2979375] [Citation(s) in RCA: 7] [Impact Index Per Article: 2.3] [Reference Citation Analysis] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Indexed: 11/07/2022]
|
12
|
Abstract
AbstractIncreasingly, people must interact with robot technologies. In this research, we examined attitudes toward robots as equipment and as coworkers and whether these attitudes are affected by the autonomy of the robot among participants living in the United States (Study 1: N = 1003; Study 2: N = 969). Study 1 revealed that respondents had a more positive attitude toward robots as equipment than as coworkers. Technology use self-efficacy and prior robot use experience were associated with more positive attitudes toward both robot positions. Having a degree in engineering or technology was associated with a positive attitude toward robot coworkers, while neuroticism was associated with a negative attitude. Additionally, technology use self-efficacy was found to have a significant indirect effect on the associations between openness and attitudes toward robots as well as conscientiousness and attitudes toward robots. In Study 2, a three-group online survey experiment showed that teleoperated robots and semi-autonomous robots were preferred as equipment over fully autonomous robots. The robots’ autonomy level did not impact attitude toward robot coworkers. Overall, the results suggest that people prefer non-autonomous robots over autonomous robots in the work-life context. The studies provide a comprehensive overview of attitudes toward robots as both equipment and coworkers, and the key predictors of the noted attitudes. The results suggest a readiness for shared autonomy between a human operator and a robot. This should be considered in the design and successful implementation of new robot technologies in workplaces.
Collapse
|
13
|
Awad LN, Lewek MD, Kesar TM, Franz JR, Bowden MG. These legs were made for propulsion: advancing the diagnosis and treatment of post-stroke propulsion deficits. J Neuroeng Rehabil 2020; 17:139. [PMID: 33087137 PMCID: PMC7579929 DOI: 10.1186/s12984-020-00747-6] [Citation(s) in RCA: 30] [Impact Index Per Article: 7.5] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Grants] [Track Full Text] [Download PDF] [Figures] [Journal Information] [Subscribe] [Scholar Register] [Received: 01/09/2020] [Accepted: 08/19/2020] [Indexed: 12/29/2022] Open
Abstract
Advances in medical diagnosis and treatment have facilitated the emergence of precision medicine. In contrast, locomotor rehabilitation for individuals with acquired neuromotor injuries remains limited by the dearth of (i) diagnostic approaches that can identify the specific neuromuscular, biomechanical, and clinical deficits underlying impaired locomotion and (ii) evidence-based, targeted treatments. In particular, impaired propulsion by the paretic limb is a major contributor to walking-related disability after stroke; however, few interventions have been able to target deficits in propulsion effectively and in a manner that reduces walking disability. Indeed, the weakness and impaired control that is characteristic of post-stroke hemiparesis leads to heterogeneous deficits that impair paretic propulsion and contribute to a slow, metabolically-expensive, and unstable gait. Current rehabilitation paradigms emphasize the rapid attainment of walking independence, not the restoration of normal propulsion function. Although walking independence is an important goal for stroke survivors, independence achieved via compensatory strategies may prevent the recovery of propulsion needed for the fast, economical, and stable gait that is characteristic of healthy bipedal locomotion. We posit that post-stroke rehabilitation should aim to promote independent walking, in part, through the acquisition of enhanced propulsion. In this expert review, we present the biomechanical and functional consequences of post-stroke propulsion deficits, review advances in our understanding of the nature of post-stroke propulsion impairment, and discuss emerging diagnostic and treatment approaches that have the potential to facilitate new rehabilitation paradigms targeting propulsion restoration.
Collapse
Affiliation(s)
- Louis N Awad
- College of Health and Rehabilitation Sciences: Sargent College, Boston University, Boston, MA, USA.
| | - Michael D Lewek
- Division of Physical Therapy, Department of Allied Health Sciences, University of North Carolina at Chapel Hill, Chapel Hill, NC, USA
| | - Trisha M Kesar
- Division of Physical Therapy, Emory University, Atlanta, GA, USA
| | - Jason R Franz
- Joint Department of Biomedical Engineering, University of North Carolina at Chapel Hill and North Carolina State University, Chapel Hill, NC, USA
| | - Mark G Bowden
- Division of Physical Therapy, Medical University of South Carolina, Charleston, SC, USA
| |
Collapse
|
14
|
Perelman BS, Evans III AW, Schaefer KE. Where Do You Think You're Going? ACM TRANSACTIONS ON HUMAN-ROBOT INTERACTION 2020. [DOI: 10.1145/3385008] [Citation(s) in RCA: 1] [Impact Index Per Article: 0.3] [Reference Citation Analysis] [Abstract] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Indexed: 10/24/2022]
Abstract
Route planning is a critical behavior for human-intelligent agent (H-IA) team mobility. The scientific community has made major advances in improving route planner optimality and speed. However, human factors, such as the ability to predict and understand teammates’ actions and goals, are necessary for trust development in H-IA teams. Trust is especially critical when agents’ behaviors do not match human team members’ expectations, or the human cannot understand the agent's underlying reasoning process. To address this issue, the artificial intelligence community has pushed toward creating
human-like
agent behaviors using machine learning. The problem with this approach is that we do not yet have a clear understanding of what constitutes human-like behavior across the breadth of tasks that H-IA teams undertake. This article describes an investigation and comparison of human and agent route planning behaviors, the interplay between humans and agents in collaborative planning, and the role of trust in this collaborative process. Finally, we propose a data-driven methodology for characterizing and visualizing differences among routes planned by humans and agents. This methodology provides a means to advance compatible mental model metrics and theory by informing targeted transparency manipulations, thereby improving the speed and quality of routes produced by H-IA teams.
Collapse
|
15
|
Rossi S, Rossi A, Dautenhahn K. The Secret Life of Robots: Perspectives and Challenges for Robot’s Behaviours During Non-interactive Tasks. Int J Soc Robot 2020. [DOI: 10.1007/s12369-020-00650-z] [Citation(s) in RCA: 8] [Impact Index Per Article: 2.0] [Reference Citation Analysis] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Indexed: 12/15/2022]
|
16
|
Sánchez López JD, Cambil Martín J, Villegas Calvo M, Luque Martínez F. [Artificial intelligence and robothics. Reflections about the need of a new bioethics framework implementation]. J Healthc Qual Res 2020; 36:113-114. [PMID: 31956011 DOI: 10.1016/j.jhqr.2019.07.009] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 07/11/2019] [Accepted: 07/18/2019] [Indexed: 10/25/2022]
Affiliation(s)
- J D Sánchez López
- Facultativo especialista de Área de Cirugía Oral y Maxilofacial, vocal del Comité Ético de Investigación de Granada, España.
| | - J Cambil Martín
- Enfermero. profesor del Departamento de Enfermería, Facultad de Ciencias de la Salud, Universidad de Granada, España
| | - M Villegas Calvo
- Enfermera, supervisora de Enfermería, H.U. Virgen de las Nieves de Granada, España
| | - F Luque Martínez
- Doctor en Farmacia, responsable de Formación, H.U. Virgen de las Nieves de Granada, vicepresidente del Comité Ético de Investigación de Granada, España
| |
Collapse
|
17
|
|