EL 03 - FlexiNav - A Ring-Based Finger-Controlled Cursor Navigator System

FlexiNav is a cutting-edge "Ring-Based Finger-Controlled Cursor Navigator System," revolutionizing assistive technology for those with Parkinson’s disease and similar tremor-inducing conditions. Addressing the challenges these conditions pose to motor functions and computer use, FlexiNav accurately translates 3D finger movements into precise cursor control on computer screens. This innovative, ring-based device not only enhances digital accessibility but also restores independence, enabling smoother navigation and interaction with digital environments for daily computer tasks. 

Our motivation stems from a deep understanding of the difficulties faced by individuals with Parkinson’s disease and other neuromotor conditions. Traditional computer mice or touchpads do not cater to their unique needs, as these devices require a level of motor control that is often compromised by the disease. To bridge this gap, we leveraged extensive data on Parkinson’s-related tremors to create a system that translates the complex 3D finger movements into precise cursor control, thus overcoming the limitations imposed by involuntary tremors.

Technical Innovation

The cornerstone of our project is its ability to discern between voluntary finger movements and involuntary tremors. This is achieved through a sophisticated combination of hardware and software that works in concert to filter out tremor-induced movements, ensuring that only intentional cursor navigations and clicks are executed.

System Components and Operation

This innovative system is meticulously engineered to provide seamless wireless operation, integrating three main components: a transmitter, a highly responsive receiver, and a computer running advanced, tailor-made software. 

Transmitter Design:

• Construction and Design: The transmitter, a pivotal component of the system, is a composite unit consisting of a main transmitting module and a specialized accelerometer ring.
• Ergonomic Placement: The design ensures ergonomic comfort, with the main unit fastening securely around the user's wrist. Simultaneously, the lightweight accelerometer ring fits securely on a finger, ideally the index or middle finger, for optimal motion capture.
• Microcontroller and Sensory Accuracy: Central to the transmitter's functionality is the Cortex M0+ RP2040 microcontroller, a choice driven by its processing power and efficiency in handling complex data streams. Complementing this is the MPU6050 gyroscope/accelerometer, a sensor meticulously chosen for its fast-sampling rate in capturing multidimensional movement data across the X, Y, and Z axes, crucial for the detailed analysis of finger motion. Lastly, the transmitter is equipped with a USB Type-C port, enabling easy connections to power banks for extended use and convenient recharging.

Receiver Design:

• Microcontroller Integration: Mirroring the transmitter's sophistication, the receiver is embedded with a Cortex M0+ RP2040 microcontroller, paired with an nRF24L01+ module, ensuring robust and uninterrupted communication with the transmitter. 
• Connectivity and Power Supply: The receiver harnesses power through a USB Type-C cable, a standard in modern electronics, which also doubles as the primary data conduit to the computer. This dual functionality not only simplifies the setup but also ensures a clutter-free user environment.
• Data Refinement and Processing: A standout feature of the receiver is its ability to process and refine incoming data meticulously. It employs a low-pass filter, fine-tuned to eliminate extraneous movements and noise, thus ensuring that only relevant motion data is transmitted to the computer software. This precision filtering is key in providing an accurate and responsive user experience.
• Interactive User Interface: The receiver boasts an advanced OLED display, complete with a comprehensive graphical user interface. This interface is intuitively designed, allowing users to effortlessly adjust crucial parameters such as sensitivity settings and nRF communication configurations. The interface's user-friendly nature makes it accessible to a wide range of users, regardless of their technical expertise.

Computer-End Software Capabilities:

• Data Reception: At the computer end, the proprietary software plays a crucial role in the system's functionality. It receives data via a Communication (COM) port connection, analyzing and interpreting crucial parameters such as the angular tilt and the frequency of tremors, a feature particularly beneficial for users with Parkinson's disease. 
• Advanced Filtering Mechanism: The software employs a sophisticated Fast Fourier Transform (FFT) filter, which serves as the first line of defense in distinguishing involuntary tremors from voluntary finger movements. This filtration process is meticulously conducted across all three axes, ensuring comprehensive data analysis.
• AI-Assisted Data Interpolation: Post-FFT, the data undergoes further refinement through an AI-assisted dynamic filter. This stage is crucial for smoothing and interpolating the data points, facilitating fluid and natural mouse movement, similar to traditional mouse devices but with enhanced precision and responsiveness.
• Linear Transformation for Cursor Navigation: The system then employs a linear transformation matrix, translating the complex 3D finger movements into 2D cursor movements on the computer screen.
• Auto-Click Functionality: A remarkable feature of the software is its ability to discern a user's intention to perform a left-click. When the auto-click function is activated on the receiver node, if the cursor is idle for a predetermined period (e.g. five seconds) over a specific pixel differential zone (e.g. 10 pixels in radius), the software performs a left click.
• Debug and Calibration Mode: To ensure optimal performance, the software includes a debug mode. This mode allows users to visually plot and monitor all incoming information and spatial data from the receiver node. It is useful for troubleshooting and calibration purposes.

Developmental Process

Our development process was methodical and phased. We began with creating a functional prototype using basic components to test the concept. Upon successful initial trials, we progressed to designing custom printed circuit boards (PCBs) for both the transmitter and receiver. This advancement was not only a technical upgrade but also enhanced the system's ergonomics, making it more user-friendly.

In parallel, our team worked on coding the system in a mixture of C/C++, Python, and MicroPython. A significant milestone was the transition from using Bluetooth to the nRF24L01+ modules for communication. This shift was driven by the need for more reliable, continuous, and faster data transmission which is a critical aspect for the system's functionality.

Testing and Results

Our testing was comprehensive, encompassing various stages of the system's development. We compared the performance of our system with traditional computer mice, focusing on tracing accuracy and responsiveness. The test requires the user to trace a given ellipse on the display once with a regular mouse and once while wearing the device.