TL;DR
This paper introduces an innovative gaze-controlled power wheelchair system designed for individuals with upper limb limitations. By leveraging an extended Function-Behavior-Structure (FBS) model and the MoSCoW method for requirement prioritization, the system offers an intuitive and safe alternative to traditional joystick controls, demonstrating promising usability and user satisfaction in initial trials.
What Problem Does This Paper Address?
Operating conventional power wheelchairs can be a significant challenge for individuals with upper limb disadvantages. Traditional joysticks require fine motor control that may be difficult or impossible for these users, leading to frustration, reduced independence, and limited participation in social activities. While various alternative control methods exist (voice, BCI, etc.), there's a continuous need for intuitive, reliable, and user-centered solutions that address specific user needs and integrate seamlessly into daily life. This paper aims to fill this gap by developing and evaluating a gaze-controlled system that prioritizes user requirements and offers a natural, efficient, and safe mode of mobility.
Methods
The core of this research lies in its systematic approach to design and evaluation.
- Extended FBS Model: The authors employed an extended Function-Behavior-Structure (FBS) model to systematically identify and define user requirements. This framework helps translate user needs into functional and structural requirements for the system.
- MoSCoW Method: To prioritize the identified requirements, the MoSCoW (Must Have, Should Have, Could Have, Won't Have) method was utilized. This ensured that the most critical functionalities were addressed first, guiding the design and development process.
- Gaze Control System: The developed system uses an eye-tracking module (Pupil Core headset) to detect the user's gaze point in the environment. This gaze information is processed through a three-layer architecture:
- Perception Layer: Acquires environmental and user eye movement data.
- Decision-Making Layer: Interprets gaze data to determine the intended direction.
- Execution Layer: Controls the wheelchair's motors based on the decision layer's output.
- Multimodal Feedback: The system incorporates audio and vibration modules to provide real-time feedback on the wheelchair's status (e.g., wake-up, stop, arrival), enhancing user awareness and safety.
- Touch Sensor for Wake-up/Emergency Stop: A touch sensor is integrated for initiating the system (wake-up) and for an immediate emergency stop, adding a crucial safety layer.
- UWB Indoor Positioning: Ultra-Wideband technology was used for precise tracking of the wheelchair's position during experimental trials to measure driving accuracy.
Datasets
The study did not utilize pre-existing datasets for training. Instead, it relied on:
- Literature Research: To understand existing user needs and challenges in wheelchair use.
- User Interviews: Two long-term power wheelchair users were interviewed to gather firsthand insights into their experiences and expectations.
- Experimental Data: Data was collected from 14 participants during simulated driving tests on a controlled experimental site. This included driving times, drift measurements, and System Usability Scale (SUS) scores.
Key Insights and Ideas
- User-Centric Design is Paramount: The study strongly emphasizes the importance of a user-centered design process, starting from understanding user needs (via literature and interviews) and systematically translating them into functional requirements using frameworks like FBS and MoSCoW.
- Gaze as an Intuitive Interface: The research validates gaze as a natural and intuitive input method for wheelchair control, particularly for users with manual control limitations.
- Layered Architecture for Robustness: The three-layer (perception, decision, execution) system architecture provides a structured and modular approach, allowing for clear separation of concerns and easier debugging/improvement.
- Multimodal Feedback Enhances Safety and Usability: The integration of both audio and vibration feedback is crucial for informing the user about the wheelchair's state, contributing to a sense of security and reducing confusion.
- Prioritization is Key for Feasible Development: The MoSCoW method proved effective in focusing development efforts on essential functionalities, ensuring a viable product within practical constraints.
Key Results
- Driving Accuracy: Participants exhibited an average driving drift of less than 20 cm, with the long path showing slightly better accuracy (average drift of 15.72 cm) than the short path (average drift of 19.49 cm). This suggests good control precision.
- Usability: The System Usability Scale (SUS) scores averaged 78.21, which is considered "good usability" according to established benchmarks. The majority of participants scored above the 70-point threshold.
- User Satisfaction: The satisfaction questionnaire revealed a high average score of 4.36 out of 5, indicating that participants were generally satisfied with the system and willing to recommend it.
- Qualitative Feedback: Participants found the system easy to understand and learn, and it effectively improved their mobility. However, some suggested improvements like reducing delay time and enhancing ride comfort.
Limitations and Future Work
- Limitations:
- Healthy Volunteers: The study used healthy volunteers with temporarily bound hands due to COVID-19 concerns and the product's early stage. This may not fully represent the experience of users with permanent disabilities.
- Indoor Environment: The experiments were conducted solely indoors, limiting the assessment of the system's performance and user experience in diverse outdoor environments.
- Preliminary Application: The current study represents a preliminary application of the extended FBS model, with further optimization and integration needed.
- Laptop Interference: The presence of a laptop for calibration and testing sometimes interfered with users' natural gaze behavior, requiring them to extend their heads.
- Future Work:
- System Stability and Timeliness: Optimize system stability, reduce delay times, and enhance overall timeliness.
- Interaction Mode Optimization: Further refine the interaction mode to reduce the cognitive and physical load on the user's head and eyes.
- Enhanced Feedback: Improve the feedback system for different states (starting, turning) to increase clarity and a sense of security.
- Outdoor Testing: Evaluate the system in real-world outdoor environments.
- Involve Users with Disabilities: Conduct experiments with individuals who have actual upper limb mobility limitations.
- Address Laptop Interference: Optimize positioning or explore alternative methods to mitigate laptop interference with gaze behavior.
Summary updated