WRLKit: Computational Design of Personalized Wearable Robotic Limbs

Paper Title

WRLKit: Computational Design of Personalized Wearable Robotic Limbs

Paper Information

• Subject Area: Wearable robotics, computer-aided design, human-computer interaction technologies
• Keywords: wearable robotics, supernumerary robotic limbs, computational design, design tools, rapid prototyping, manufacturing

Research Background and Problem

• Research Problem: Wearable robotic limbs (WRL) are technologies that attach to the human body and enhance human capabilities. However, current designs primarily target robotics experts, making it difficult for ordinary users and non-professional designers to participate in the customization and development of these devices. Additionally, existing design methods are time-consuming and complex, failing to meet individual and dynamic human needs.
• Research Significance: Expanding wearable robotic design from expert domains to ordinary non-professional users, enabling them to quickly design personalized devices on demand. This is particularly relevant for human-machine collaboration, novel interaction design, and customized functional development.
• Related Work:
  • Existing studies mostly focus on WRLs for specific tasks or body parts but lack flexibility.
  • Related literature has explored the interaction between body dynamics and robotic device control, but design tools have yet to integrate the design and manufacturing process.
  • Current computational design and manufacturing methods (e.g., robotic optimization design, sensor personalization) provide inspiration for the development of this system.

Solution

• Proposed Method: WRLKit is an interactive computational design tool driven by human body data. It employs optimization algorithms to achieve personalized modeling of dynamic human characteristics and rapid prototyping of devices.
• Innovations:
  1. Introducing task-based dynamic posture capture for the first time to enable customized wearable robotic design.
  2. Integrating customizable target area settings (e.g., specifying body surfaces or surrounding spatial regions).
  3. Automatically generating modular robotic arm structures with 3 degrees of freedom (DoF).
  4. Considering multiple objectives, including human parameters, range of motion, comfort of use, device portability, and collision avoidance.
• Implementation Steps:
  1. Capture and Model Human Information: Using RGB cameras and MoCap technology to obtain user body dimensions and dynamic postures.
  2. Set Design Parameters: Users select device placement and target space (body or spatial regions) via a graphical user interface.
  3. Optimize Device Structure: The system generates robotic arm designs that meet personalized needs through multi-objective optimization.
  4. Design Review and Iteration: Users can validate the design and further optimize it by adjusting parameters.
  5. Generate Manufacturing Files: Export design results into files suitable for 3D printing and laser cutting.

Research Outcomes

• Specific Results:
  1. Developed a human-centered computational design tool, WRLKit.
  2. Achieved real-time capture of human postures and generation of robotic limb configurations.
  3. Successfully prototyped various WRLs for different applications, such as:
     • Upper-arm-mounted WRL providing haptic feedback (for VR use).
     • Waist-mounted WRL assisting with button pressing.
     • Forearm-mounted WRL for object stabilization.
  4. Verified the tool's applicability for non-professional users through experiments.
• Advantages:
  1. Deep integration of dynamic human motion and robotic design, significantly reducing design iteration time.
  2. User-friendly interface, making complex technologies (e.g., robotic kinematics optimization) transparent to users.
  3. Modular design facilitates rapid assembly and reuse.
• Experiments and Evaluation:
  1. In user experiments, participants generally mastered the tool's logic quickly and completed device designs within 7 minutes.
  2. Users acknowledged the tool's ease of use and speed, eliminating the need for manual engineering design.
• Limitations and Future Directions:
  1. Current designs are limited by the 3 DoF robotic arm structure, restricting supported task types.
  2. Dynamic behavior models of WRLs (e.g., real-time load control) are not fully considered.
  3. The detail of body surface modeling can be further improved.
  4. Users expressed a desire for more flexibility in interaction design, such as manual adjustment of different joint types.
  5. In the future, the system could integrate AR/VR technologies for real-time scene simulation and design visualization.

Conclusion

WRLKit is a revolutionary interactive design tool that makes wearable robotic technology more accessible to designers outside the robotics domain. By combining dynamic human posture modeling, rapid optimization algorithms, and modular manufacturing methods, the system demonstrates significant potential in the rapid development of personalized devices.