Design Process of Rigid and Soft Robotic Gloves for Hand Rehabilitation Combining Clinical Research Methodologies and Engineering Practice

Design Process of Rigid and Soft Robotic Gloves for Hand Rehabilitation Combining Clinical Research Methodologies and Engineering Practice Mihai Dragusanu Department of Information Engineering and Mathematics University of Siena Siena, Italy 0000-0001-9373-6866 Monica Malvezzi Department of Information Engineering and Mathematics University of Siena Siena, Italy 0000-0002-2158-5920 Abstract—Designing effective hand rehabilitation devices requires integrating clinical needs with engineering solutions. This study shows how to apply the PICO framework (Population/Problem, Intervention, Comparison, Outcome) to guide the development of two robotic hand exoskeletons: a rigid device and a tendon-driven soft glove. The rigid exoskeleton provides precise, repeatable finger actuation, suitable for patients requiring controlled movements, while the soft glove offers compliant, adaptable assistance for users with low hand strength or spasticity. Comparative analysis highlights differences in actuation, adaptability, and user interaction. Preliminary tests demonstrate improved range of motion and exercise repeatability for both devices. The PICO approach proves a practical tool to systematically translate clinical requirements into user-centered rehabilitation device design. Index Terms—exoskeletons, rehabilitation robotics, wearable robots I. INTRODUCTION The development of rehabilitative technologies increasingly requires methods that place the user at the center of the design process. In both industry and healthcare, there has been a clear shift from technology-driven solutions toward approaches that prioritize usability, adaptability, and patient-specific needs [1]– [4]. This shift reflects a broader recognition that technological innovation alone is not sufficient unless it directly addresses the requirements of those who use the devices. The rehabilitation field exemplifies this challenge: patients present highly diverse clinical conditions, levels of impairment, and recovery trajectories, which demand devices that are not only functional but also adjustable to individual anatomical and functional characteristics. Current trends in engineering address this need through the development of lightweight materials, ergonomic structures, embedded sensors, and advanced control strategies. Furthermore, methods such as 3D scanning and the use of customizable geometries allow devices to be tailored more closely to patient anatomy, thereby improving comfort, usability, and therapeutic effectiveness [5], [6]. We gratefully acknowledge the support of the European Union by the Next Generation EU project ECS00000017 “Ecosistema dell’Innovazione” Tuscany Health Ecosystem (THE, PNRR: Spoke 9: Robotics and Automation for Health) Despite these advances, a persistent challenge remains: how to systematically translate clinical requirements into concrete design parameters. Clinicians typically define problems in terms of functional impairments, therapeutic interventions, and expected outcomes, whereas engineers rely on performance metrics, technical specifications, and design constraints. This misalignment often complicates the process of transforming clinical insights into engineering solutions. Without a structured method to bridge these perspectives, there is a risk of developing devices that are technologically sophisticated but misaligned with user needs, or conversely, clinically relevant concepts that lack technical feasibility. Establishing a common framework that supports communication and knowledge transfer across disciplines is therefore essential to ensure that rehabilitation technologies are both clinically relevant and technically robust. One promising avenue for addressing this gap is the transfer of methodologies from clinical research into design practice [7]. The PICO framework (Population/Problem, Intervention, Comparison, Outcome) is widely employed in Evidence-Based Medicine (EBM) to structure clinical questions and guide the systematic evaluation of interventions. Its value lies in its ability to distill complex problems into well-defined elements that can be analyzed and compared in a consistent way. Although rarely considered in engineering, PICO’s structured, outcome-oriented logic offers an opportunity to enrich the design process. By framing design problems around users, interventions, alternatives, and desired results, PICO can function as a practical tool within a user-centered design (UCD) approach. In doing so, it supports the integration of clinical insights into engineering practice, while also encouraging a systematic evaluation of alternative design options. This paper explores the use of PICO as a design aid for robotic rehabilitation devices. We argue that adopting PICO within the design process can strengthen interdisciplinary communication, reduce information loss between clinical and engineering domains, and provide a clear pathway from medical needs to technical solutions. More broadly, PICO can be viewed as a methodological bridge between evidence-based 2025 I-RIM Conference October 17-19, Rome, Italy ISBN: 9788894580570 10.5281/zenodo.17629866 225 reasoning and user-centered design, aligning device development more closely with patient outcomes. To demonstrate this potential, we extend the study presented in [8] and propose a comparative analysis of two hand exoskeletons: one based on rigid-link mechanisms [9]–[11] and another designed as a soft-glove [12]–[14]. This case study illustrates how PICO can be adapted and applied in practice, ultimately supporting engineers in creating rehabilitation technologies that are effective and responsive to the diverse needs of users. II. DESIGN GUIDED BY THE PICO FRAMEWORK A. Introduction to the PICO framework The PICO framework is traditionally used to formulate research questions and guide meta-analyses and clinical studies [7]. In rehabilitation, it can also serve as a tool to organize and clarify the design of assistive and robotic devices. Built around four elements, the PICO provides a structured method for translating clinical needs into engineering requirements. The P defines the target users and their impairments, such as stroke survivors, post-surgical patients, or individuals with neurological disorders [7]. The I specifies the therapeutic strategy, which may involve exoskeletons, prosthetics, or digital rehabilitation platforms, and helps identify relevant functional improvements and technical constraints [15]. The C evaluates the proposed solution against existing treatments or devices, offering insights into performance, regulatory validation, and potential advantages [16]. Finally, O captures the measurable results, such as functional recovery, usability, and user satisfaction, that both guide design choices and serve as benchmarks for effectiveness. In this way, PICO supports the systematic identification of clinical priorities and their translation into technical features, offering designers and researchers a practical framework for developing rehabilitation devices that are evidence-based and user-centered. Placing the patient at the center of the design process is now widely recognized as best practice in medical device development, as end-user characteristics directly inform the functional and technical specifications of a device. Many of these considerations are also captured by the PICO framework, although from a clinical perspective. In the following, we present two case studies that illustrate how design constraints derived from PICO can be systematically translated into engineering requirements and applied directly in the development of new prototypes. B. The rigid hand Exoskeleton A hand exoskeleton for flexion/extension assessment and training was developed with a user-centered approach, allowing autonomous donning, modularity, and use in both therapist-assisted and independent rehabilitation [9]–[11], [17]. The final version comprises five finger modules actuated jointly or independently, integrating FMA MicroForce sensors (Honeywell, North Carolina) to measure forces applied by the patient or therapist. The CAD model is shown in Fig. 1a while in Fig. 1b the device is worn by and end-user. (a) (b) Fig. 1: Rigid Exoskeleton. (a) CAD model. (b) The hand exoskeleton worn by an end-user. The device includes a Hand Support and modular actuated fingers. The support, shaped to fit the hand, combines internal cushioning with external ABS-CF10 carbon fiber and is secured via TPU straps. Finger modules are mounted on a horizontal slot with a Swing Pivot Point allowing flexion/extension and lateral/longitudinal adjustment. The thumb uses a distinct strap-based mechanism (Fig. 1b) to accommodate its unique kinematics. Actuation is provided by PQ12-63-6-R linear actuators (Actuonix), with two for the thumb and one per other finger, controlled via Teensy 4.1 controller. Force sensors are housed within the actuator supports, with TPU Distributors on the thumb ensuring safety and measurement accuracy. Each phalanx support includes a soft TPU half-ring for secure attachment. The mechanical design follows postural synergies [18], aligning linkage dimensions and positions with the first synergy of each finger. The exoskeleton allows flexion/extension for all fingers, while abduction/adduction is implemented only passively, except for the thumb. The design and development can be framed by the PICO approach. a) Population/Problem (P): This rigid exoskeleton is designed as a rehabilitation and training tool for patients with limited hand control, low residual force and mild spasticity. According to the impairment/pathology mapping in [7], these limitations may arise from neurological conditions such as stroke. Unlike soft devices, the rigid structure provides precise actuation and force transmission, making it suitable for 226 patients who require controlled, repeatable finger movements rather than high adaptability or flexibility. b) Intervention (I): The exoskeleton supports grasping and releasing objects while guiding rehabilitation exercises. It allows the recording and repetition of predefined motion profiles via a user interface. Each finger module provides one degree of freedom for active flexion/extension, while the thumb also enables adduction/abduction. The rigid design ensures accurate force application and consistent motion trajectories, critical for effective therapy, though the number of actively actuated motions per finger is limited. c) Comparison (C): While no quantitative comparison with alternative therapies has been performed, the device has been evaluated alongside other rigid robotic hands in terms of actuation type, mechanical structure, number of actuated fingers, tip force, system weight, portability, motion type, and modularity [9]. This analysis highlights the advantages of a rigid exoskeleton in delivering precise, controlled, and repeatable rehabilitation exercises. d) Outcome (O): Preliminary tests with potential users indicate repeatability and consistency of rehabilitative exercises. Motion tracking is achieved through integrated sensors, including IMU-based systems adapted from earlier designs [9], providing real-time feedback via a graphical user interface. The modular structure allows selective use of modules (e.g., thumb and index), and parametric parts can be manufactured using standard FDM technologies, making the device adaptable to different users and rehabilitation needs. C. The MGlove-TS The design of an actuated glove for hand rehabilitation following an impairmentguided approach is presented in Dragusanu et al. 2024 [12] (see Fig. 2). The device is a tendon-driven soft glove actuated by twisted string actuators (TSA), designed with a parametric and modular approach to accommodate different users’ needs. Each finger module can function independently and may incorporate up to two tendons per finger. In the configuration shown in Fig. 2b, the thumb and index modules are worn by the user, with tendons arranged to enable flexion, adduction, and abduction motions. Finger extension can be achieved either passively via an elastic structure on the back of the fingers or actively using a dual-TSA arrangement, as illustrated in Fig. 2a. Tendons are routed through adjustable anchor points on the glove in contact with the palm and connected to an actuation unit located on the forearm. The effect of tendon routing and anchor placement on finger motion control has been analyzed in [14], [19], comparing oneand two-tendon finger module designs. TSA actuators can be embedded in a forearm-mounted bracelet using small motors for portability and comfort. Also in this case, the design and development can be framed by the PICO approach. a) Population/Problem (P): The P addressed by this glove includes patients with low active hand control, reduced residual force, low muscle tone, and mild spasticity, potentially (a) (b) Fig. 2: Rigid Exoskeleton. (a) CAD model. (b) The hand exoskeleton worn by an end-user. resulting from neurological conditions such as Parkinson’s disease or Limb-Girdle Muscular Dystrophy [7]. For these patients, soft devices are preferred over rigid or hybrid solutions due to their high adaptability and flexibility, while the force and precision requirements remain moderate. b) Intervention (I): The I provided by the glove supports grasping and releasing objects while guiding rehabilitation exercises. The 2-DoF finger modules described in [12] enable flexion, adduction, and abduction motions, and are controlled via a graphical user interface (GUI) with LeapMotion tracking [13]. This setup allows the recording and repetition of predefined motion profiles. c) Comparison (C): Concerning C no quantitative evaluation against alternative therapies has been performed. However, the glove has been analyzed relative to other robotic gloves developed over the past decade in terms of actuation, mechanical transmission, number of actuated fingers, tip force, system weight, portability, motion type, modularity, and intrinsic safety [12]. This comparison highlights the advantages of the soft glove in terms of adaptability, comfort, and flexibility. For the Outcome (O), preliminary tests with potential users demonstrate improvements in joint range of motion (ROM). For example, index finger adduction/abduction increased from 8° without assistance to 15° while wearing the glove. Integrated sensors, including IMU-based systems adapted from previous wearable designs [9], [13] or tracking 227 system as LeapMotion [13] provide real-time feedback via the GUI, guiding exercise execution through visual or auditory cues. The modular structure allows selective use of finger modules (e.g., thumb and index only), and parametric parts can be manufactured using standard FDM technologies, enabling customization for different users. ROM limits were set within physiological ranges to ensure safety, and the glove is lightweight, flexible, and designed to minimize interference with natural hand movements. A removable motor and TSA pin system ensures safety in the event of motor failure or blockage, maintaining back-drivability and preventing the hand from being locked in a flexed position. III. DISCUSSION Applying the PICO framework provided a structured approach to guide the design of both the rigid and soft hand exoskeletons. For P, the rigid exoskeleton targets patients with low hand control but who require precise, repeatable finger movements, whereas the soft glove is better suited for patients needing flexible, adaptable assistance due to low force and mild spasticity. In terms of I, both devices support grasping and releasing objects and facilitate rehabilitation exercises, but the rigid exoskeleton emphasizes controlled, active actuation of all fingers, while the glove leverages tendon-driven soft modules for more compliant and adaptable motion. Regarding C, analysis against alternative devices and therapies highlighted the advantages of the rigid exoskeleton in delivering repeatable, high-precision movements, while the glove excels in adaptability, comfort, and ease of use for varied patient anatomies. Finally, the O of preliminary tests demonstrate that both systems can improve joint range of motion and exercise repeatability, with integrated sensors providing feedback and monitoring. The rigid device ensures precise force control, whereas the soft glove offers safer, more flexible interactions with the user’s hand. IV. CONCLUSION The PICO framework proved effective in systematically translating clinical needs into engineering requirements for both rigid and soft hand exoskeletons. It allowed clear identification of user characteristics, appropriate interventions, relevant comparisons, and measurable outcomes, guiding the design process and ensuring that the resulting devices meet patient-specific rehabilitation goals. The rigid exoskeleton is particularly suitable for patients requiring precise and repeatable movements, while the soft glove provides adaptable and compliant assistance for users with lower residual strength. Overall, PICO offers a practical methodology to integrate clinical insight into the user-centered design of robotic rehabilitation devices, supporting both technical robustness and patient safety. REFERENCES [1] B. ¨ Ostlund, M. Malvezzi, S. Frennert, M. Funk, J. Gonzalez-Vargas, K. Baur, D. Alimisis, F. Thorsteinsson, A. Alonso-Cepeda, G. Fau et al., “Interactive robots for health in europe: Technology readiness and adoption potential,” Frontiers in Public Health, vol. 11, p. 979225, 2023. [2] M. Dragusanu, Design of Soft–Rigid Devices for Rehabilitative and Assistive Robotics. Springer Nature, 2025. [3] S. Frennert, J. Persson, and S. Skavron, “A critical narrative review of assistive robotics and call for a systems and user-centered approaches to enhance quality of life of individuals with disabilities,” in Adjunct Proceedings of the 2024 Nordic Conference on Human-Computer Interaction, 2024, pp. 1–11. [4] S. Helmstetter, M. D¨ orr, R. Germann, and S. Matthiesen, “User-centered design of power tools: a generic process for evaluation of usability aspects,” Forschung im Ingenieurwesen, vol. 86, no. 1, pp. 93–104, 2022. [5] C. G. Rose and M. K. O’Malley, “Hybrid rigid-soft hand exoskeleton to assist functional dexterity,” IEEE Robotics and Automation Letters, vol. 4, no. 1, pp. 73–80, 2018. [6] J. Arata, K. Ohmoto, R. Gassert, O. Lambercy, H. Fujimoto, and I. Wada, “A new hand exoskeleton device for rehabilitation using a three-layered sliding spring mechanism,” in 2013 IEEE international conference on robotics and automation. IEEE, 2013, pp. 3902–3907. [7] G. M. Achilli, C. Amici, M. Dragusanu, M. Gobbo, S. Logozzo, M. Malvezzi, M. Tiboni, and M. C. Valigi, “Soft, rigid, and hybrid robotic exoskeletons for hand rehabilitation: roadmap with impairmentoriented rationale for devices design and selection,” Applied Sciences, vol. 13, no. 20, p. 11287, 2023. [8] C. Amici, R. Buraschi, M. Dragusanu, M. Gobbo, S. Logozzo, M. Malvezzi, J. Pollet, M. Tiboni, and M. C. Valigi, “Beyond boundaries: Pico-driven design criteria for robotic rehabilitation medical devices,” in International Workshop IFToMM for Sustainable Development Goals. Springer, 2025, pp. 285–292. [9] M. Dragusanu, M. Z. Iqbal, T. L. Baldi, D. Prattichizzo, and M. Malvezzi, “Design, development, and control of a hand/wrist exoskeleton for rehabilitation and training,” IEEE Transactions on Robotics, vol. 38, no. 3, pp. 1472–1488, 2022. [10] M. Dragusanu, D. Troisi, D. Prattichizzo, and M. Malvezzi, “Evaluating adaptability properties of a compact wearable hand exoskeleton,” in 2024 10th IEEE RAS/EMBS International Conference for Biomedical Robotics and Biomechatronics (BioRob), 2024, pp. 1561–1566. [11] M. Dragusanu, A. Piroli, and M. Malvezzi, “Human–exoskeleton interaction force measurement in a wearable device for hand actuation,” in International Workshop on Medical and Service Robots. Springer, 2025, pp. 251–260. [12] M. Dragusanu, D. Troisi, B. Suthar, I. Hussain, D. Prattichizzo, and M. Malvezzi, “Mglove-ts: A modular soft glove based on twisted string actuators and flexible structures,” Mechatronics, vol. 98, p. 103141, 2024. [13] M. Dragusanu, D. Troisi, B. Suthar, D. Prattichizzo, and M. Malvezzi, “Development of a soft actuated glove based on twisted string actuators for hand rehabilitation,” in 2024 10th IEEE RAS/EMBS International Conference for Biomedical Robotics and Biomechatronics (BioRob), 2024, pp. 1702–1708. [14] M. Dragusanu, A. Saeed, N. Guinet, D. Troisi, D. Prattichizzo, and M. Malvezzi, “Development of the modular finger elements of an actuated glove for hand rehabilitation,” in The International Conference of IFToMM ITALY. Springer, 2024, pp. 563–570. [15] W. S. Richardson, M. C. Wilson, J. Nishikawa, and R. S. Hayward, “The well-built clinical question: a key to evidence-based decisions.” ACP journal club, vol. 123, no. 3, pp. A12–3, 1995. [16] R. Formicola, C. Amici, M. Mor, L. Bissolotti, and A. Borboni, “Design of medical devices with usability in mind: a theoretical proposal and experimental case study using the lepre device,” Designs, vol. 7, no. 1, p. 9, 2023. [17] Design of a Modular Hand Exoskeleton for Rehabilitation and Training, ser. ASME International Mechanical Engineering Congress and Exposition, vol. Volume 5: Biomedical and Biotechnology, 11 2021. [Online]. Available: https://doi.org/10.1115/IMECE2021-70343 [18] M. Santello, M. Flanders, and J. F. Soechting, “Postural hand synergies for tool use,” Journal of neuroscience, vol. 18, no. 23, pp. 10 105–10 115, 1998. [19] M. Dragusanu, A. Piroli, and M. Malvezzi, “Analysis and development of the modular elements of a tendon-actuated glove for hand rehabilitation,” Robotica, p. 1–19, 2025. 228