Jun Nishida 西田纯University of Maryland 马里兰大学College Park, Maryland, United States 美国马里兰州学院帕克jun@umd.edu
Figure 1: (a) We developed a hand thermal I/O exoskeleton, DexteriSync, that can dynamically adjust finger temperature. (b) This can computationally modulate a user's finger dexterity, which we validated through a pegboard evaluation. (c) As such, it can, for instance, allow a product designer to experience the lived experience of having compromised dexterity seen in neurodivergent children or the elderly, resulting in helping designers design efficient tools through embodied product evaluation. 图 1:(a)我们开发的手部热输入/输出外骨骼 DexteriSync 可以动态调节手指温度。(b) 这可以通过计算调节用户的手指灵活性,我们通过钉板评估对此进行了验证。(c)因此,它可以让产品设计师体验到神经变异儿童或老人灵巧性受损的生活经历,从而通过体现式产品评估帮助设计师设计出高效的工具。
Abstract 摘要
Skin temperature is an important physiological factor for human hand dexterity. Leveraging this feature, we engineered an exoskeleton, called DexteriSync, that can dynamically adjust the user's finger dexterity and induce different thermal perceptions by modulat ing finger skin temperature. This exoskeleton comprises flexible silicone-copper tube segments, 3D-printed finger sockets, a 3D printed palm base, a pump system, and a water temperature control with a storage unit. By realising an embodied experience of compromised dexterity, DexteriSync can help product designers understand the lived experience of compromised hand dexterity, such as that of the elderly and/or neurodivergent users, when designing daily necessities for them. We validated DexteriSync via a technical evaluation and two user studies, demonstrating that it can change skin temperature, 皮肤温度是影响人类手部灵活性的重要生理因素。利用这一特点,我们设计了一种名为 DexteriSync 的外骨骼,它可以动态调整用户的手指灵活性,并通过调节手指皮肤温度诱发不同的热感。这种外骨骼由柔性硅铜管段、3D 打印手指套、3D 打印手掌底座、泵系统和带存储单元的水温控制器组成。DexteriSync 实现了手部灵活性受损的具身体验,有助于产品设计师在为老年人和/或神经变异用户设计日常用品时,了解他们手部灵活性受损的生活体验。我们通过技术评估和两项用户研究对 DexteriSync 进行了验证,证明它可以改变皮肤温度、
dexterity, and thermal perception. An exploratory session with design students and an autistic compromised dexterity individual, demonstrated the exoskeleton provided a more realistic experience compared to video education, and allowed them to gain higher confidence in their designs. The results advocated for the efficacy of experiencing embodied compromised finger dexterity, which can promote an understanding of the related physical challenges and lead to a more persuasive design for assistive tools. 灵巧性和热感知。与设计专业的学生和一名自闭症灵巧性受损者进行的探索性课程表明,与视频教育相比,外骨骼提供了更真实的体验,让他们对自己的设计更有信心。结果证明了体验手指灵活性受损的效果,这可以促进对相关身体挑战的理解,从而设计出更有说服力的辅助工具。
hand exoskeleton, dexterity, thermal perception, embodied interaction 手部外骨骼、灵巧性、热感知、体现式交互
ACM Reference Format: ACM 参考格式:
Ximing Shen, Youichi Kamiyama, Kouta Minamizawa, and Jun Nishida. 2024. DexteriSync: A Hand Thermal I/O Exoskeleton for Morphing Finger Dexterity Experience. In The 37th Annual ACM Symposium on User Interface Software and Technology (UIST '24), October 13-16, 2024, Pittsburgh, PA, USA. ACM, New York, NY, USA, 12 pages. https://doi.org/10.1145/3654777.3676422 沈锡明、神山友一、南泽幸太和西田淳。2024.DexteriSync:用于变形手指灵巧体验的手部热输入/输出外骨骼。第 37 届 ACM 用户界面软件与技术研讨会(UIST '24),2024 年 10 月 13-16 日,美国宾夕法尼亚州匹兹堡。ACM, New York, NY, USA, 12 pages.https://doi.org/10.1145/3654777.3676422
1 INTRODUCTION 1 引言
The human hand is a primary media to interact with the surrounding environment, attributed to its provision of dexterous hand interactions, tactile feedback, and multi-touch capabilities [4, 25, 77, 87]. Finger dexterity is one of the crucial factors as it enables everyday fine tasks including writing, shirt-buttoning, and object-grasping [84]. Human-Computer Interaction (HCI) researchers have proposed interaction techniques to increase or convey the normal level of hand dexterity, such as increasing dexterity by Electrical Muscle Stimulation (EMS) [67, 90], restoring manual dexterity for stroke patients using active exoskeletons [2, 95], and even communicating one's dexterous skills between two users using a paired exoskeleton [66]. These approaches allowed users to enhance their motor skills, which are otherwise hard to communicate and acquire in visual contact, and further gain their embodied knowledge faster and more accurately. 人的手是与周围环境进行交互的主要媒介,因为它具有灵巧的手部交互、触觉反馈和多点触控功能[4, 25, 77, 87]。手指的灵巧性是关键因素之一,因为它可以完成日常精细任务,包括书写、扣衬衫纽扣和抓取物体 [84]。人机交互(HCI)研究人员提出了一些交互技术来提高或传递手部的正常灵巧程度,例如通过肌肉电刺激(EMS)提高灵巧性[67, 90],使用主动外骨骼恢复中风患者的手部灵巧性[2, 95],甚至使用配对外骨骼在两个用户之间交流一个人的灵巧技能[66]。这些方法使用户能够提高他们的运动技能,而这些技能很难在视觉接触中交流和获得,并进一步更快、更准确地获得他们的具身知识。
Meanwhile, less is known about reproducing compromised dexterity, and how such disabling experience can contribute to practical scenarios. Owing to the impaired interaction between sensory and motor functions, compromised dexterity happens to a large number of the population, such as elderly or people with neurological disorders including Autism Spectrum Disorder (ASD) [15], forearm paralysis caused by cerebral palsy [17], and mild cognitive impairment [21]. 与此同时,人们对再现失灵的灵活性以及这种失灵体验如何有助于实际场景的了解却较少。由于感觉功能和运动功能之间的相互作用受损,很多人都会出现灵活性受损的情况,如老年人或患有自闭症(ASD)[15]、脑瘫引起的前臂瘫痪[17]和轻度认知障碍[21]等神经系统疾病的人。
Fewer works have looked into techniques to replicate the everyday challenges faced by users who have decreased dexterity to help surrounding users (e.g., nursing/medical staff, and product designers) understand their needs, the nature of symptoms, and associated subjective experiences. For example, taping the user's thumb and forefinger of the dominant hand to simulate the loss of hand motor skills [70], or restricting each finger's moving range with a glove to simulate the feeling of having arthritis . While these methods can mimic the physical performance of having constrained hand movement range, because multiple physiological and neurological factors influence dexterity, they (1) do not capture the full experience of having compromised dexterity, which is therefore difficult to argue for the amount of the information provided that is supposed to lead to practical solutions for designers and medical professionals; (2) have limited adjustability, which restricts the range of conditions and scenarios that they can represent. 较少作品研究复制灵活性下降的用户所面临的日常挑战的技术,以帮助周围的用户(如护理/医疗人员和产品设计师)了解他们的需求、症状的性质以及相关的主观体验。例如,用胶带绑住用户惯用手的拇指和食指来模拟手部运动技能的丧失[70],或用手套限制每个手指的活动范围来模拟患有关节炎的感觉 。虽然这些方法可以模拟手部活动范围受限的物理表现,但由于多种生理和神经因素会影响灵巧性,它们(1)无法捕捉灵巧性受损的完整体验,因此很难论证所提供的信息量是否能为设计师和医疗专业人员提供实用的解决方案;(2)可调整性有限,这限制了它们所能表现的条件和场景的范围。
For these reasons, there is still the need to: (1) design new wearable techniques to reproduce the compromised dexterity through the intervention towards its physiological mechanisms to provide a further informative experience, (2) provide the sensory experience related to motor performance, such as changes in thermal perception that can be seen in specific user groups including ASD [11] and Multiple Sclerosis (MS) [20], and (3) explore how the embodied experience of having compromised dexterity can support daily interactions, such as providing new inspirations during the product design process. 由于这些原因,仍有必要(1)设计新的可穿戴技术,通过对其生理机制的干预来重现受损的灵活性,从而提供进一步的信息体验;(2)提供与运动表现相关的感官体验,例如在特定用户群体(包括 ASD [11] 和多发性硬化症(MS)[20])中可以看到的热感知变化;以及(3)探索灵活性受损的体现体验如何支持日常互动,例如在产品设计过程中提供新的灵感。
1.1 Our Approach 1.1 我们的方法
Hence, in this paper, we developed a hand thermal I/O exoskeleton, called DexteriSync, to provide a further realistic and computationally adjustable experience of having compromised dexterity. Our 因此,在本文中,我们开发了一种名为 DexteriSync 的手部热输入/输出外骨骼,以提供更逼真的、可通过计算调整的灵活性受损体验。我们的
approach is inspired by the nature of human thermal regulation in its interconnection with tactile sensation [30] and motor performance [24]. We explored the functionality of the exoskeleton in the following two aspects: 这种方法的灵感来源于人类热调节与触觉[30]和运动表现[24]之间的相互联系。我们从以下两个方面探索了外骨骼的功能:
1.1.1 Morphing Manual Dexterity. Temperature can change blood flow [85], tendon flexibility [71] and muscle contraction [6], which all contribute to the mobility of the human hand. Because these three factors play pivotal roles in regulating skin temperature, hand dexterity has been found positively related to human skin temperature [13]. More specifically, cold exposure is correlated with decreased fine motor performance, in the dimension of dexterity [8], pinch [12] and grip force [14]. Inspired by this phenomenon, we explored whether the user's dexterity can be altered if we use our exoskeleton to deliver cold temperature changes to the user's finger skin in User Study 1. 1.1.1 手部灵活性的变形。温度可以改变血流量[85]、肌腱柔韧性[71]和肌肉收缩[6],这些因素都有助于提高人手的灵活性。由于这三个因素在调节皮肤温度方面起着关键作用,因此人们发现手的灵活性与人体皮肤温度呈正相关[13]。更具体地说,在灵巧性[8]、捏力[12]和握力[14]方面,暴露于寒冷环境与精细运动能力下降相关。受这一现象的启发,我们在用户研究 1 中探索了如果使用我们的外骨骼为用户的手指皮肤提供低温变化,是否可以改变用户的灵活性。
1.1.2 Morphing Thermal Perception. Thermal perception is a decisive factor in behaviours and decision-making [82], likewise motor performance [7]. Past works have shown that making alterations to human skin can induce abnormal thermal perceptual phenomena . We adopted this approach and investigated whether imbalanced finger temperature would lead to a different thermal perception if we use the exoskeleton to deliver independent temperature to a user's fingers simultaneously in User Study 2. 1.1.2 变形热感知。热感知是行为和决策的决定性因素 [82],同样也是运动表现的决定性因素 [7]。过去的研究表明,改变人体皮肤可诱发异常的热感知现象 。我们采用了这种方法,并在用户研究 2 中调查了如果我们使用外骨骼同时为用户的手指提供独立的温度,手指温度的不平衡是否会导致不同的热感知。
Exploratory Session. We also conducted an exploratory session, which involved design students and a user with ASD and compromised dexterity, to assess the effectiveness of using the exoskeleton to gain embodied knowledge of compromised hand dexterity while designing assistive attachments for the target user. 探索环节。我们还进行了一次由设计专业学生和一名患有自闭症且手部灵活性受限的用户参与的探索环节,以评估在为目标用户设计辅助附件时,使用外骨骼获取手部灵活性受限的体现性知识的有效性。
2 RELATED WORKS 2 部相关作品
2.1 Dexterity and Skin Temperature 2.1 灵巧度和皮肤温度
Dexterity is the fine movements of the hands and fingers that one uses to perform tasks. From simple tasks such as shoe-tying and shirt-buttoning, to digital activities that involve tapping and swiping on the smartphone, plenty of daily activities require dexterous interactions [28, 84, 86]. Dexterity is affected by many factors, notably age [55], finger thickness [104], and skin temperature of the hand [8]. Among these, skin temperature is a key variable [73] that can influence dexterity by altering sensory perception [99], blood flow [8], and muscle contraction [98]. For this reason, cold environments can reduce workers' hand dexterity [75]; while extreme temperatures ( e.g., outdoor tennis) can decrease athletes' performance [93]. Following this path, prior works relied on the relationship between skin temperature and dexterity to investigate optimal room temperature for precision-required workplaces (e.g., surgery rooms) [32], cold exposure work guidelines [75], sports training strategies [29, 93], and rehabilitation therapies [89, 97]. Research of such sheds light on the potential of using skin temperature to explore hand functions in future interaction design. 灵巧是指人在执行任务时双手和手指的精细动作。从系鞋带、扣衬衫纽扣等简单的任务,到在智能手机上轻点和轻扫的数字活动,大量的日常活动都需要灵巧的互动[28, 84, 86]。灵活性受多种因素影响,主要有年龄[55]、手指粗细[104]和手部皮肤温度[8]。其中,皮肤温度是一个关键变量[73],可通过改变感觉[99]、血流[8]和肌肉收缩[98]来影响灵活性。因此,寒冷环境会降低工人的手部灵活性[75];而极端温度(如室外网球)会降低运动员的表现[93]。按照这一思路,之前的研究依赖于皮肤温度和灵活性之间的关系,研究了精密要求工作场所(如手术室)的最佳室温[32]、寒冷暴露工作指南[75]、运动训练策略[29, 93]和康复疗法[89, 97]。这些研究揭示了在未来的交互设计中利用皮肤温度探索手部功能的潜力。
2.2 Interactive Devices to Change Hand Functions 2.2 改变手部功能的互动设备
In HCI, one popular way to explore hand functionality is to harness visual perception, and numerous Virtual Reality (VR) research 在人机交互领域,探索手部功能的一种常用方法是利用视觉感知,而众多虚拟现实(VR)研究
Figure 2: Details of (a) silicone-copper tube segments, (b) a finger socket, (c) a palm base, and (d) the exoskeleton configuration. 图 2:(a) 硅铜管段、(b) 指套、(c) 手掌底座和 (d) 外骨骼结构的细节。
studies were centred around this idea. Researchers have found that changing the perceived hand size can affect task efficiency [50], perceived skin tone can induce embodiment [56], and hand temper ature colour can influence the actual hand temperature [44]. This has been extended to wearable devices, such as wearing soft robotic fingers for prosthetic augmentation [36] or a passive exoskeleton for grasp range miniaturisation [63]. On the other side, dexterity has been explored via different digital means. EMS has been explored as it has the potential to address dexterous movements [53], with added brakes [67] or dorsal stimulation [90]. Others attempted to increase dexterity using exoskeletons with DC motors [78] or pneumatic actuators or even communicate dexterous movements between two users [66] using a passive exoskeleton. These previous works demonstrated the feasibility of using mechanical actuators to increase hand functions and dexterity. Meanwhile, we opted for a thermal haptic approach because (1) finger temperature is a major factor influencing dexterity [8], and (2) abnormal thermal perception also contributes to a more realistic compromised experience seen in the hand disorder population [35]. 研究都围绕这一观点展开。研究人员发现,改变感知到的手的大小可影响任务效率[50],感知到的肤色可诱导体现[56],手的温度颜色可影响实际手温[44]。这一点已延伸到可穿戴设备,如佩戴软机器人手指用于假肢增强[36]或被动外骨骼用于抓取范围微型化[63]。另一方面,人们通过不同的数字手段探索灵巧性。EMS 具有解决灵巧运动的潜力[53],并增加了制动[67]或背侧刺激[90]。还有人尝试使用带有直流电机[78]或气动致动器 的外骨骼来提高灵巧性,甚至使用被动外骨骼在两名用户之间交流灵巧动作[66]。这些前人的工作证明了使用机械致动器提高手部功能和灵活性的可行性。与此同时,我们选择了热触觉方法,因为(1)手指温度是影响灵巧性的主要因素[8],(2)异常的热感知也有助于手部疾病患者获得更真实的受损体验[35]。
2.3 Interactive Thermal Research in HCI 2.3 人机交互中的交互热研究
There is increasing research exploring thermal stimuli as an HCI modality. Peltier modules are commonly used to create temperature feedback, and have been put to use on the wrist [68], forearm [16, 52], finger , auricle [61], tongue , or multiple body parts [62, 102]. However, they face limitations, namely the narrow temperature range, short operation time, the significant power consumption, and the bulkiness of heatsinks and cooling fan for heat dissipation. Liquid cooling is an alternative heat transfer method effective at heat dissipation. It can handle high heat loads, and has flexibility in terms of design, allowing configure a thermal display combined with resistive heating [46, 47] or a cooling wearable system using cold water and a Peltier module [72, 74]. However, they also have limitations such as slower temperature changes (due to water's high heat capacity) and additional setup including the water tank. 将热刺激作为人机交互模式的研究越来越多。珀尔帖模块通常用于产生温度反馈,并已应用于手腕 [68]、前臂 [16, 52]、手指 、耳廓 [61]、舌头 或身体多个部位 [62, 102]。然而,它们面临着一些限制,即温度范围窄、工作时间短、功耗大,以及散热片和冷却风扇体积大。液体冷却是另一种有效的热传递方法。它能处理高热负荷,设计灵活,可配置与电阻加热相结合的热显示屏 [46, 47] 或使用冷水和珀尔帖模块的可穿戴冷却系统 [72, 74]。不过,它们也有局限性,如温度变化较慢(由于水的热容量大)和包括水箱在内的额外设置。
On the other side, researchers found that thermal stimuli can influence people's emotional response [1]. For this reason, numerous thermal feedback research probed how it can be utilised for VR immersiveness. These include presenting thermal feedback to haptic gloves [40, 42, 49], headsets [69], an upper-body [31], or a hand controller [34]. Meanwhile, dexterity remains an untapped area in HCI thermal research. Leveraging the characteristic of skin temperature affecting motor performance, we provide our design and evaluations of the novel thermal I/O hand exoskeleton that affects dexterity. 另一方面,研究人员发现热刺激可以影响人们的情绪反应[1]。因此,许多热反馈研究都在探索如何利用热反馈来增强虚拟现实的沉浸感。这些研究包括向触觉手套[40, 42, 49]、头盔[69]、上半身[31]或手部控制器[34]提供热反馈。同时,灵巧性仍然是人机交互热研究中一个尚未开发的领域。利用皮肤温度影响运动性能的特性,我们设计并评估了可影响灵活性的新型热敏 I/O 手部外骨骼。
2.4 Embodied Simulations of Bodily Difference 2.4 身体差异的具身模拟
At the same time, researchers are also exploring how body-altering technologies can be used to understand physical differences through users' embodied experiences. One popular approach is perspectivetaking activities in Augmented Reality (AR) and VR. For example, using AR to simulate children's visible height [64] or visual impairments of people with dementia [83], using VR to simulate first-person-view of different skin tone [56] or the everyday office experience of neurodiversity worker [54]. Another approach is the embodied hand experience of physical challenges using interactive wearable devices. Such as using passive exoskeletons to simulate children's hand size [63] or the restricted function of arthritis , and using EMS to simulate the tremor of Parkinson's disease [65]. Besides these single-channel experiences, there are also multisensory simulation toolkits of ASD and the elderly [27]. Such simulations are now used to foster accessibility education [83] and enhance universal design [63]. Inspired by these approaches, we set to explore how the compromised thermal and dexterity experience can enhance the embodied product design procedures and provide new insights into this topic. 与此同时,研究人员还在探索如何利用改变身体的技术,通过用户的具身体验来了解身体差异。一种流行的方法是在增强现实(AR)和虚拟现实(VR)中开展透视活动。例如,使用 AR 模拟儿童的可见身高 [64] 或痴呆症患者的视觉障碍 [83],使用 VR 模拟不同肤色的第一人称视角 [56] 或神经多样性工作者的日常办公体验 [54]。另一种方法是利用交互式可穿戴设备来体验体能挑战。例如使用无源外骨骼模拟儿童手的大小[63]或关节炎的功能受限 ,以及使用 EMS 模拟帕金森病的震颤[65]。除了这些单通道体验外,还有 ASD 和老年人 的多感官模拟工具包 [27]。这类模拟现在被用于促进无障碍教育 [83] 和加强通用设计 [63]。受这些方法的启发,我们开始探索受损的热感和灵巧体验如何能够增强体现式产品设计程序,并为这一主题提供新的见解。
3 IMPLEMENTATION 3 实施
In this section, we describe the design details of our exoskeleton prototype, which is depicted in Figure 3. It consists of four components: 1) flexible silicone-copper tube segments, 2) 3D-printed finger sockets and palm base, 3) a pump system, and 4) a water temperature control with a storage unit. 在本节中,我们将介绍外骨骼原型的设计细节,如图 3 所示。它由四个部分组成:1)柔性硅铜管段;2)3D 打印指套和手掌底座;3)泵系统;4)带存储单元的水温控制装置。
The part attached to the finger was inspired by multiple personal thermal management methods [79], in which we used a hybrid cooling method combining material conduction and water cooling convection to address several challenges that existing implementations faced. In particular: (1) it exhibits robust thermal performance in reflecting changes to the skin temperature of the human finger; (2) each finger unit combined materials weigh about 6.2 g , hence 附在手指上的部分受到多种个人热管理方法的启发[79],其中我们使用了一种结合材料传导和水冷对流的混合冷却方法,以解决现有实现方法面临的几个挑战。特别是:(1) 它在反映人体手指皮肤温度变化方面表现出稳定的散热性能;(2) 每个手指单元的材料组合重量约为 6.2 克,因此,它的散热性能较好。
Figure 3: Overall configuration of the DexteriSync prototype. 图 3:DexteriSync 原型的整体配置。
the lightweight does not add burden to the user's hand; (3) the soft material covering the dorsal side does not interfere with the palm side activities or moving range of the hand/arm; (4) its length can be flexibly adjusted by adding/removing silicone tubes and thus, is adjustable for users of all hand/arm size. (3) 背侧覆盖的柔软材料不会影响手掌侧的活动或手/臂的活动范围;(4) 其长度可通过增减硅胶管灵活调节,因此适合各种手/臂尺寸的使用者。
3.1 Exoskeleton Prototype 3.1 外骨骼原型
Flexible silicone-copper tube segments. We covered the finger skin on the dorsal side using copper tubes for heat conductivity, and the knuckles with silicone tubes for flexible finger movement. We chose copper for its efficient heat conductivity (approximately at room temperature). We cut (LDT) copper pipes (uxcell) into 25 mm long pieces using a Proxxon MICRO band saw MBS 240/E with a band saw blade No. 28174, which is suitable for cutting thin metals. We then used a metal hand file to smooth out the rough cut. The finished piece is between and hammered into an oval shape, maximising the contact area with the skin of the finger. We cut ( ) silicone tubes (uxcell) into 20 mm pieces for the knuckles, and connected the two materials using Cemedine Super X adhesive During the user studies, the length of the silicone tube extended to the arm part was adjusted using ( ) PVC connectors (uxcell) to accommodate each participant's hand and arm size. 灵活的硅胶-铜管片段。我们用铜管覆盖手指背侧的皮肤,以利于导热,用硅胶管覆盖手指关节,以利于手指灵活运动。我们选择铜,是因为它具有高效的导热性(室温下约为 )。我们使用 Proxxon MICRO 带锯 MBS 240/E,将 (LDT) 铜管 (uxcell)切割成 25 毫米长的小块,带锯条编号为 28174,适用于切割薄金属。然后,我们用金属手锉将粗糙的切口打磨平整。成品在 之间,锤成椭圆形,最大限度地增加了与手指皮肤的接触面积。我们将 ( ) 硅胶管(uxcell)切割成 20 毫米的小块,用于指关节,并使用 Cemedine Super X 粘合剂将两种材料连接起来 在用户研究期间,我们使用 ( ) PVC 连接器(uxcell)调整硅胶管延伸至手臂部分的长度,以适应每位参与者的手和手臂尺寸。
3D-printed finger sockets and palm base. We designed the finger socket, which serves two purposes: 1) the two bottom pieces help to attach the silicone-copper tube segments onto the finger skin, and 2) the middle upper piece holds the folded-back silicone tubes in place. The palm base serves one purpose: to fix all tube segments onto the back of the hand. All 3D-printed finger sockets and the palm base were manufactured using black vero material in a Stratasys J55 3D printer, and were further dyed using black matte spray paint. 3D 打印指套和手掌底座。我们设计的指套有两个作用:1)底部两片帮助将硅铜管段固定在手指皮肤上;2)中间上片将折回的硅胶管固定到位。手掌底座的作用只有一个:将所有管段固定在手背上。所有 3D 打印手指套和手掌底座都是使用黑色维罗材料 在 Stratasys J55 3D 打印机上制造的,并使用黑色哑光喷漆进行了进一步染色。
3.2 Cooling System 3.2 冷却系统
A pump system. We opted for water cooling because of water's high heat capacity . We connected each extended silicone tube of the exoskeleton to one tube pump PP310-030. To maximise the water flow, we used three pumps for each finger, and the pump setup for each finger is controlled by a Locomo board . 泵系统。我们选择水冷却,因为水的热容量 很高。我们将外骨骼的每根延长硅胶管连接到一个管泵 PP310-030。为了最大限度地提高水流量,我们为每个手指使用了三个泵,每个手指的泵设置由一块 Locomo 电路板 控制。
A water temperature control and storage unit. We modified a thermoelectric assembly by switching to M4x20 mm screws to let the pump tubes pull through. The tubes were sandwiched between the Peltier modules and the heatsinks. We used a K-type thermocouple and a MAX6675 thermocouple amplifier to monitor the water temperature inside the cooler box at all times. The Peltier modules were controlled by a Proportional, Integral, Derivative (PID) controller using an Arduino Uno to provide real-time temperature control. We used one control unit when keeping the temperatures of the two fingers identical, and two control units when controlling the finger temperatures individually. 水温控制和储存装置。我们改装了一个热电组件,改用 M4x20 毫米的螺钉让泵管穿过。泵管夹在珀尔帖模块和散热器之间。我们使用 K 型热电偶和 MAX6675 热电偶放大器随时监测冷却器箱内的水温。珀尔帖模块由使用 Arduino Uno 的比例、积分、微分(PID)控制器控制,以提供实时温度控制。当两个手指的温度保持一致时,我们使用一个控制单元;当单独控制手指温度时,我们使用两个控制单元。
We adopted the following formula to use time of wearing the exoskeleton to influence the skin temperature of the finger based on Newton's law of cooling: . We kept the ambient temperature at degree, the contact area of skin and the exoskeleton at approximately , and the water temperature at . Meanwhile, the heat capacity of human skin is between and based on [100], mass of the skin and cooling coefficient differ by each individual's skin thickness and individual variation. 根据牛顿冷却定律,我们采用以下公式利用穿戴外骨骼的时间 来影响手指皮肤温度 : 。我们将环境温度 保持在 度,皮肤与外骨骼的接触面积 约为 ,水温为 。同时,根据[100],人体皮肤的热容量 在 和 之间,皮肤的质量 和冷却系数 因每个人的皮肤厚度和个体差异而不同。
Figure 4: Time-series skin temperature profiles. 图 4:时间序列皮肤温度曲线。
3.3 Technical Evaluation 3.3 技术评估
Purpose and method: Heat transfer can exhibit a time delay due to factors including thermal mass and thermal conductivity of objects. Therefore, we measure a time-series profile of skin temperature changes by attaching a K-type thermocouple on the palm side finger front ( 6 males, 6 females; years old, ). All experiments started at the same hour of the day ( 1 PM ) under the same room temperature , participants wore about the same amount of clothing (short/non-sleeve T-shirts) and verbally confirmed that their hands were at normal state before the experiment. We chose for the cold water and for the hot water based on Fourier's Law of Heat Conduction: . When the thermal conductivity of the material and the area of the heat being transferred are the same, maximising the initial temperature difference can lead to a higher heat transfer rate . Therefore, we chose the water temperatures that can create large temperature differences from the skin temperature while not getting frostbite (0 or burnt . 目的和方法:由于物体的热质量和热传导率等因素,热传导会出现时间延迟。因此,我们在手掌侧手指前端(6 男 6 女; 岁, )安装 K 型热电偶,测量皮肤温度变化的时间序列剖面图。所有实验都在同一时间(下午 1 点)开始,室温 相同,参与者穿着差不多的衣服(短袖/无袖 T 恤),并在实验前口头确认双手处于正常状态。根据傅立叶热传导定律,我们选择 为冷水, 为热水: 。当材料的导热系数 和传热面积 相同时,最大化初始温差 可带来更高的传热速率 。因此,我们选择了既能与皮肤温度产生较大温差,又不会冻伤(0 或烧伤 )的水温。
Summary: Figure 4 shows our time-series skin temperature profile. Under the cold conditions, it took about 10 minutes until the 总结:图 4 显示了皮肤温度的时间序列曲线。在寒冷条件下,大约 10 分钟后,皮肤温度才会升高。
Figure 5: (a) Experiment setup used in the dexterity study. (b) An example of the pegboard pattern. 图 5: (a) 灵巧性研究中使用的实验装置。(b) 挂板图案示例。
skin temperature of the thumb reached and became stable. It took about 10 minutes until the skin temperature of the index finger reached and became stable. Under the hot conditions, it took about 15 minutes until the skin temperature of the thumb reached and became stable. It took about 15 minutes until the skin temperature of the index finger reached and became stable. 拇指的皮肤温度达到 并趋于稳定。大约 10 分钟后,食指的皮肤温度达到 并趋于稳定。在高温条件下,约 15 分钟后,拇指的皮肤温度才达到 并趋于稳定。大约 15 分钟后,食指的皮肤温度达到 并趋于稳定。
4 USER STUDIES 4 用户研究
To understand how our exoskeleton modulates a wearer's motor and sensory experience, we investigated the changes in the following two factors: 1) hand dexterity and 2) thermal perception in a lab setting. Our user studies were approved by our institutional review board. 为了了解我们的外骨骼如何调节穿戴者的运动和感官体验,我们研究了以下两个因素的变化:1)手部灵活性;2)实验室环境下的热感。我们的用户研究获得了机构审查委员会的批准。
4.1 User Study 1: Dexterity Changes 4.1 用户研究 1:灵活性的变化
The objective of this study is to understand how DexteriSync decreases a user's hand dexterity. We adopted a within-subject approach to investigate the participants' dexterity performance under multiple time conditions. Our study was a dexterity experiment modified from the standardised Purdue Pegboard Test [94]. 本研究旨在了解 DexteriSync 如何降低用户的手部灵活性。我们采用受试者内部研究法,调查受试者在多种时间条件下的灵巧性表现。我们的研究是根据标准化的普渡钉板测试(Purdue Pegboard Test)[94] 修改而成的灵巧性实验。
Our priori hypothesis is that if DexteriSync can decrease a participant's finger skin temperature, then their dexterity can be decreased This prediction was grounded in earlier physiology research studies that cold exposure can decrease hand performance [8, 13, 57]. However, in these papers, participants were asked to put their arms into cold water or wear a cooling vest, which changed the temperature of broader body parts and that of the core body temperature. In our study, we evaluated the sole effect of finger skin temperature on hand dexterity. 我们的先验假设是,如果 DexteriSync 可以降低参与者的手指皮肤温度,那么他们的灵活性就会下降。这一预测的依据是早期的生理学研究,即暴露在寒冷环境中会降低手部表现 [8、13、57]。然而,在这些论文中,参与者被要求将手臂放入冷水中或穿上降温背心,这就改变了身体更广泛部位的温度和核心体温。在我们的研究中,我们只评估了手指皮肤温度对手部灵活性的影响。
4.1.1 Task and Setup. In each trial, participants were shown a different pegboard pattern with the same level of difficulty ( 10 pegs, same colour, three times of direction change). Participants were asked to place the pegs into holes following the pattern and colour. The pegboard patterns were randomised to avoid the learning effect An experimenter measured the time spent using their smartphone. 4.1.1 任务和设置。在每次试验中,参与者都会看到一个难度相同的不同钉板图案(10 个钉子,相同颜色,三次方向变化)。参与者被要求按照图案和颜色将钉子放入孔中。为了避免学习效应,钉板图案是随机的。实验人员测量了参与者使用智能手机的时间。
We kept the cold water at degree throughout the study. There was no direct contact between the participants' skin and the water we used. 在整个研究过程中,我们将冷水的温度保持在 度。参与者的皮肤与我们使用的水没有直接接触。
4.1.2 Condition. Participants wore DexteriSync on the dominant hand and performed the task under the following four conditions: The cooling system was turned off, and no temperature changes happened (baseline); and then the cooling system ran for 2) 5 minutes, 3) 10 minutes, or 4) 15 minutes to cool down the finger; 4.1.2 条件。受试者将 DexteriSync 佩戴在惯用手上,在以下四种条件下执行任务:冷却系统关闭,没有温度变化(基线);然后冷却系统运行 2) 5 分钟、3) 10 分钟或 4) 15 分钟,为手指降温;
4.1.3 Procedure. Each participant conducted the experiment from 1 PM , the participant sat in a chair while exposed to a room temperature of degree [8]. In total, each participant performed a total of 4 trials. At the end of each condition, participants were asked to answer two 7-point Likert scale questions regarding their sense of control (Statement 1) and ownership (Statement 2). These questions are described in Figure 7 and were designed based on existing sense of agency and ownership questionnaires [5, 9, 92]. 4.1.3 实验过程每位受试者从下午 1 点开始进行实验,受试者坐在椅子上,室温为 度[8]。每位被试总共进行了 4 次试验。在每个条件结束时,受试者被要求回答两个有关其控制感(陈述 1)和自主感(陈述 2)的 7 分李克特量表问题。这些问题如图 7 所示,是根据现有的代理感和所有权问卷设计的[5, 9, 92]。
4.1.4 Participants. We recruited 6 healthy right-handed participants ( 2 female, 3 male, 1 prefer not to answer, years old, ) from our institution through a snowball sampling method. Participants performed the study for about 60 minutes, and received 20 USD as compensation for their time. 4.1.4 参与者。我们通过 "滚雪球 "的方法在本机构招募了 6 名健康的右利手参与者(2 名女性,3 名男性,1 名不愿回答, 岁, )。参与者进行了约 60 分钟的研究,并获得了 20 美元的时间补偿。
Figure 6: Participants' completion time of the pegboard task in the four conditions. 图 6:参与者在四种情况下完成钉板任务的时间。
4.1.5 Results: Completion Time. Figure 6 exhibits the completion time of pegboard tasks across the conditions. A One-Way Repeated Measures ANOVA revealed a significant main effect of the time condition . Post hoc pairwise comparisons using Fisher's LSD (see -values on Figure 6 for detail) revealed that the completion time for both condition 5 min and condition 15 min are significantly higher than that of the baseline condition ( ). This supports our main hypothesis that DexteriSync can decrease participants' dexterity when applying cold temperatures compared to the baseline condition of no temperature changes. 4.1.5 结果:完成时间图 6 显示了不同条件下钉板任务的完成时间。单向重复测量方差分析显示,时间条件 具有显著的主效应。使用费雪LSD进行事后配对比较(详见图6中的 值)发现,条件5分钟和条件15分钟的完成时间均显著高于基线条件( )。这支持了我们的主要假设,即与没有温度变化的基线条件相比,DexteriSync 可以降低参与者在使用低温时的灵活性。
4.1.6 Questionnaire Results. Figure 7 summarises our Likert scale question results. Overall, participants experienced a lower sense of control and finger ownership under the cold condition compared to the baseline condition (Statement 1: score Statement 2: score . Participants experienced the lowest sense of control at 5 min (Statement 1: score ), and lowest sense of finger ownership at 15 min (Statement 2: score=4.50, SD=2.07). 4.1.6 问卷调查结果。图 7 总结了李克特量表问题的结果。总体而言,与基线条件相比,参与者在寒冷条件下的控制感和手指所有权较低(陈述 1:得分 陈述 2:得分 )。参与者在 5 分钟时的控制感最低(陈述 1:得分 ),在 15 分钟时的手指自主感最低(陈述 2:得分=4.50,标准差=2.07)。
4.1.7 Discussion. In sum, our results suggest that our exoskeleton was effective in decreasing participants' finger dexterity under cold temperatures. The decrease-increase-decrease pattern in dexterity 4.1.7 讨论。总之,我们的结果表明,我们的外骨骼能有效降低参与者在低温条件下的手指灵活性。灵活性的下降-增加-下降模式
Figure 7: Participants' self-reported experience with regards to their sense of agency and ownership when wearing DexteriSync under the Baseline and cold conditions. 图 7:在基线和寒冷条件下,参与者自我报告佩戴 DexteriSync 时的代入感和自主感。
performance might be caused by physiological responses. The increase at 10 min might be caused by users' acclimitisation to the cold and a temporary arousal of the body's stress response [41] Meanwhile, the decrease at 15 min might be due to the cumulative effect of the cold exposure since prolonged narrowing of blood vessels might have led to a significant further decrease in muscle function . 这可能是生理反应造成的。10 分钟时的成绩提高可能是由于使用者对寒冷的适应以及身体应激反应的暂时唤醒[41],而 15 分钟时的成绩下降则可能是由于寒冷暴露的累积效应,因为血管的长时间狭窄可能会导致肌肉功能进一步显著下降 。
While participants' perceived sense of agency (Statement 1) and ownership (Statement 2) decreased under cold conditions, the changing patterns did not align with that of dexterity. We captured multiple instances in which participants believed that they had increased control over the pegboard task while their actual performance decreased. This phenomenon may have arisen from participants feeling of being acclimated to the temperature, leading them to believe that it positively impacts their performance. This further indicates that participants' acclimation to the temperature might have a bigger influence on their sense of agency than witnessing their hand dexterity. This aligns with Koreki et al. [45]'s research that interoceptive awareness is correlated to the sense of agency. 在寒冷条件下,参与者的代理感(陈述 1)和自主感(陈述 2)都有所下降,但变化模式与灵巧性的变化模式并不一致。我们多次发现,参与者认为自己对钉板任务的控制能力增强了,但实际表现却下降了。这种现象可能是由于受试者对温度产生了适应感,从而认为温度会对他们的表现产生积极影响。这进一步表明,参与者对温度的适应可能比见证他们手的灵活性对他们的代入感有更大的影响。这与 Koreki 等人[45]的研究一致,即感知间意识与代入感相关。
On the other side, participants' sense of ownership scores were close between 5 min and 10 min , which indicates a state of perceived balance. The sudden decrease in the 15 min condition suggests that there might be a threshold level between 10 min and 15 min for a significant decrease in both dexterity performance and the sense of ownership. This further suggests that controlling the time of use can be an essential safety concern for future thermal interaction design. Our findings support prior work that temperature can impact body ownership and human proprioception [37]. On top of this, we highlight that the sense of ownership is a more effective perceptual indicator of temperature-oriented physical performance compared to the sense of agency. 另一方面,参与者的主人翁感得分在 5 分钟和 10 分钟之间比较接近,这表明参与者处于一种感知平衡的状态。在 15 分钟的条件下,得分突然下降,这表明在 10 分钟和 15 分钟之间可能存在一个临界值,即灵巧性和主人翁感都会显著下降。这进一步表明,在未来的热交互设计中,控制使用时间可能是一个至关重要的安全问题。我们的研究结果支持之前的研究,即温度会影响身体的自主性和人体本体感觉[37]。此外,我们还强调,与代入感相比,自主感是以温度为导向的身体表现的更有效的感知指标。
4.2 User Study 2: Thermal Perception 4.2 用户研究 2:热感知
In this study, we explored how a user's thermal perception can be changed when experiencing imbalanced temperature between the thumb and index finger using DexteriSync. Our study was modified from a past thermal perception experiment [23], where contactless thermal radiant stimuli were used to explore the neurological effect of a thermal perceptual phenomenon, called thermal grill illusion. 在这项研究中,我们利用 DexteriSync 探索了当用户的拇指和食指之间温度不平衡时,用户的热感知会发生怎样的变化。我们的研究是在过去的热感知实验[23]基础上修改而成的,在该实验中,我们使用非接触式热辐射刺激来探索一种热感知现象的神经效应,这种现象被称为热烧烤幻觉。
Figure 8: Experiment setup and task in the thermal perception study. 图 8:热感知研究中的实验装置和任务。
We hypothesise that if DexteriSync can deliver independent temperature changes to the skin of a user's fingers simultaneously, the imbalance in temperature would lead to a disparate thermal perception of the modified fingers compared to other fingers. This hypothesis was built upon past neurology research studies that altering human skin temperature can induce thermal perceptual phenomena [22, 23, 58, 81]. However, these previous works investigated how thermal perceptual phenomena can be induced on the forearm or among the index, middle and ring fingers. In this study, we examined whether similar phenomena can be induced only between the thumb and index finger. We found a significant value in recreating the temperature imbalance because the thermal grill illusion induces an unpleasant sensation when touching objects, which is commonly experienced by the hand disorder populations . This study focuses on simulating this sensory experience of the target population, rather than the physical dexterity experience in Study 1. We believe that this sensory simulation is important for understanding their experience with products, therefore helping DexteriSync users achieve better texture and/or shape design decisions, which we further show in Section 4.3. 我们假设,如果 DexteriSync 能够同时为用户的手指皮肤提供独立的温度变化,那么温度的不平衡将导致修改后的手指与其他手指产生不同的热感知。这一假设建立在过去的神经学研究基础上,即改变人体皮肤温度会诱发热感知现象[22, 23, 58, 81]。不过,这些研究都是针对如何在前臂或食指、中指和无名指之间诱发热感知现象进行的。在本研究中,我们研究了类似现象是否只能在拇指和食指之间诱发。我们发现重现温度不平衡现象具有重要价值,因为热烧烤幻觉会在触摸物体时诱发一种不愉快的感觉,而这种感觉是手部失调人群 普遍经历过的。本研究的重点是模拟目标人群的这种感官体验,而不是研究 1 中的身体灵活性体验。我们认为,这种感官模拟对于了解他们的产品体验非常重要,因此可以帮助 DexteriSync 用户做出更好的纹理和/或形状设计决策,我们将在第 4.3 节中进一步说明这一点。
4.2.1 Task and Setup. We followed a within-subject approach to explore the participants' subjective thermal perception between their two hands' thumbs and index fingers. 4.2.1 任务和设置。我们采用受试者内部研究的方法来探究受试者双手拇指和食指之间的主观热感。
We prepared one cup of water for each hand in each trial. Participants were asked to put their thumb and index finger into the cup in front. Participants were asked to identify whether both cups of water felt the same in terms of temperature. 我们为每次试验的每只手都准备了一杯水。参与者被要求将拇指和食指放入前面的杯子中。我们要求受试者辨别两杯水的温度是否相同。
For the DexteriSync device, we kept the cold water at degree, and warm water at degree throughout the study. Across all conditions, warmth was applied to the thumb, and coldness was applied to the index finger. This was modelled after a previous finding that in a three-fingers experiment setup, placing the cold target finger in the centre (warm-cold-warm) can induce the thermal grill illusion more significantly than placing it on the side (cold-warm-warm) [58]. 对于 DexteriSync 设备,我们在整个研究过程中将冷水温度保持在 度,将温水温度保持在 度。在所有条件下,拇指都是温水,食指是冷水。这是根据以前的一项研究结果模拟的,即在三指实验设置中,将冷的目标手指放在中间(暖-冷-暖)比放在边上(冷-暖-暖)更能引起热烧烤幻觉[58]。
4.2.2 Conditions. Participants were asked to wear DexteriSync on their dominant hand and keep their non-dominant hand bare hand throughout the study. 4.2.2 条件。参与者被要求在整个研究过程中将 DexteriSync 佩戴在惯用手上,非惯用手保持裸露。
The cup of water presented to the bare hand was maintained at degree. The temperature of the cup of water presented to the DexteriSync hand was changed each time. 裸手端起的水杯温度保持在 度。DexteriSync 手上水杯的温度每次都会改变。
Following the staircase procedure methodology [19], we separated the study into two conditions. We referred to a previous study on the Just Noticeable Difference (JND) in temperature [48] to design our adjustment intervals. 按照阶梯程序方法[19],我们将研究分为两种情况。在设计调整区间时,我们参考了之前关于温度的 "刚注意到的差异"(JND)的研究[48]。
In condition low to high, we started by presenting the DexteriSync hand with cool water at degree; and increased the water temperature by degree whenever the participant identified the DexteriSync hand as noticeably colder. If the participant reported that the DexteriSync hand is noticeably warmer than the bare hand, we decreased the water temperature by degree. When the participant reported two cups felt identical, the experimenter asked about their confidence in the decision verbally. Once the participant confirmed their decision, the condition stopped and the experimenter recorded the temperatures of both cups of water. 在从低到高的条件中,我们首先用 度的凉水向 DexteriSync 手演示;每当被试认为 DexteriSync 手明显更冷时,我们就将水温提高 度。如果受试者报告 DexteriSync 手明显比裸露的手温暖,我们就将水温降低 度。当被试者报告说两个杯子的感觉相同时,实验员会口头询问他们对这一决定的信心。一旦被试确认了他们的决定,实验就会停止,实验员会记录下两杯水的温度。
In condition high to low, we started by presenting the DexteriSync hand with warm water at degree; and decreased the water temperature by degree whenever the participant identified the DexteriSync hand as noticeably warmer. If the participant reported that the DexteriSync hand is noticeably cooler than the bare hand, we increased the water temperature by degree. When the participant reported two cups felt identical, the experimenter asked about their confidence in the decision verbally. Once the participant confirmed their decision, the condition stopped and the experimenter recorded the temperatures of both cups of water. 在从高到低的条件中,我们先用 度的温水给 DexteriSync 手加热;每当被试认为 DexteriSync 手明显变热时,我们就将水温降低 度。如果受试者称 DexteriSync 手明显比裸手凉,我们就将水温升高 度。当被试者报告说两个杯子的感觉相同时,实验员会口头询问他们对这一决定的信心。一旦被试确认了他们的决定,实验就会停止,实验员会记录下两杯水的温度。
4.2.3 Procedure. Participants were asked to perform both conditions alternatively for 3 times. Only two identical cups were presented to the participants at a time. On average, participants performed the task for 18.4 times (median=19, ). 4.2.3 步骤。参与者被要求在两种条件下交替进行 3 次。每次只向受试者展示两个相同的杯子。被试平均完成任务 18.4 次(中位数=19, )。
4.2.4 Participants. We recruited eight new healthy right-handed participants ( 4 females, 4 males, years old, ) from our institution using a snowball sampling method. Participants performed the experiment for about 90 minutes, and received 20 USD as compensation for their time. 4.2.4 参与者。我们采用 "滚雪球 "的方法,在本机构招募了 8 名新的健康右利手参与者(4 名女性,4 名男性, 岁, )。参与者进行了约 90 分钟的实验,并获得了 20 美元作为时间补偿。
4.2.5 Results. Figure 9 (a) depicts the average water temperature of the bare hand and the DexteriSync hand when participants reported the same. A paired sample t-test was performed to examine the difference in thermal perception between the bare hand and the DexteriSync hand. The one-tailed t-test result revealed that the perceptual temperature of the DexteriSync hand is significantly higher than the bare hand . Figure represents the relationship between the actual water temperature and the perceived water temperature. A linear regression model was fitted for the relationship between the actual and perceived temperature of the DexteriSync hand . 4.2.5 结果图 9 (a) 显示了裸手和 DexteriSync 手的平均水温。为了检验裸手和 DexteriSync 手的热感差异,我们进行了配对样本 t 检验。单尾 t 检验结果显示,DexteriSync 手的感知温度明显高于裸手 。图 表示实际水温与感知水温之间的关系。