Access-H20 Faucet for Spinal Cord Injury

Nearly 300,000 individuals are living with spinal cord injuries (SCI), and over 12,000 new SCI occur annually in the United States. Most SCIs involve the cervical (neck) region of the spine and cause partial or complete loss of movement of both arms, both legs, and the torso (quadriplegia). Individuals with a higher-level and more complete SCI typically have greater loss of functional movement and sensation below the level of the injury. They require greater levels of assistance for self-care and mobility. Currently, there are limited options for quadriplegics to access water to perform basic activities of daily living such as drinking. Additionally, hydration packs and straws create unsanitary conditions and can be difficult to use. The aim of the project is to develop an innovative fountain faucet to enable individuals living with SCI to access water more independently for basic activities of daily living. The faucet incorporates electromechanical controls and software programs that adjust spray outlets, spray angles, and water temperatures. A proximity sensor and voice recognition function allow users to control the faucet through a smart speaker.

The Access-H2O™ faucet was built on a robust software platform, designed to integrate seamlessly with three sensor controls-eye gaze, motion, and voice control-accommodating diverse user needs (Figure 3). It features three distinct modes: fountain mode, which provides an upward fountain flow for drinking; washing mode, which delivers a traditional downward flow for handwashing and other cleaning tasks; and grooming mode, which utilizes a spray pattern for rinsing the face.

For motion control, time-of-flight sensors enable users to activate washing mode by waving or placing a hand near the proximity-based motion sensor, and then switch to fountain mode by waving again. Additionally, these sensors allow users to adjust water temperature. The software translates sensor data into actions based on three defined detection zones-hot, warm, and cold-determined by the proximity of the user's hand to the faucet. During setup, users can configure these distances to map specific ranges for hot, warm, or cold water. For example, placing a hand closest to the sensor activates hot water, while the farthest position activates cold water.

For eye gaze control, the faucet was integrated with Tobii eye-tracking technology, combining both hardware and software. To maintain a compact design without making the faucet excessively wide, we engineered a removable escutcheon to house the eye-tracking sensors. This removable feature ensured versatility, allowing the smart faucet to be installed in both 1-hole and 3-hole sink configurations. The system enables users to activate and deactivate water flow by staring at specific sensor locations for a set duration. The sensor locations are corresponded to three faucet modes. A user can stare at the right sensor to activate the fountain mode for fountain flow for drinking, the center sensor for washing mode, and the left sensor for grooming mode. The selected mode remains active until the user looks at the same sensor again to turn it off. To prevent accidental activations, the software measures fixation duration and incorporates error-handling protocols to account for distractions or environmental factors.

For voice control, users can activate three faucet modes through spoken commands.

With Alexa or Google Assistant, they can turn the faucet on or off by saying, "Alexa, turn on washing mode." Additionally, users can select specific functions, such as "Alexa, activate fountain mode" to enable the fountain feature. To control flow duration, users can issue commands like "Alexa, run water for 10 seconds."

A total of 18 subjects living with SCI (C2-C6), called SCI subjects, and 10 control subjects were recruited in an outpatient rehabilitation clinic. Controls without limitations in head, trunk, or arm movement were used to verify that all faucet functions operated correctly and that water was accessible for drinking, rinsing, and grooming. This preliminary assessment by the controls was conducted after installing the Access-H2OTM faucet in the rehabilitation clinic and before SCI subjects' participation.

The feasibility testing was conducted to test the faucet's eye gaze, voice, and motion sensors for SCI subjects' ADLs - drinking, rinsing, and grooming. We sought answers to the following questions: 1) Can the test participants get the water from the fountain mode to their mouth to drink/hydrate? 2) Can the test participants use the fountain mode to rinse their mouths? and 3) Can the test participants use the fountain spray to rinse a quarter-sized area of soap partially or fully from their left or right cheek

All SCI subjects who relied on wheelchairs were seated in their own manual or power wheelchairs. The wheelchairs were positioned in front of a sink equipped with the Access-H2OTM faucet. Sufficient space was provided to accommodate the subjects' legs under the sink, ensuring standardized data collection. Then, the same physical therapist who conducted the interview provided scripted instructions to each participant on how to turn on the faucet for the faucet modes, and then turn off the faucet. During setup, the system was tested and calibrated to ensure proper operation. If any function did not meet the standard, manual adjustments were made as needed. For example, during setup, flow and spray angles for fountain mode and grooming mode can be customized by adjusting valve angles or water pressure to achieve the preferred water flow trajectory.

Once the system was ready, each SCI subject was asked to place his or her hand by the motion sensor (motion control), use a speech speaker (voice control), and move their eyes up, down, left, and right to activate the eye gaze sensor (eye gaze control) to activate water output modes for drinking, rinsing, and grooming. For example, to test eye gaze control, a subject looked at the right side of the eye gaze sensor to activate fountain mode for drinking and looked at the left side of the sensor to activate grooming mode for spray pattern flow to wash a quarter-sized area of soap from the left or right cheek. For each ADL activity (drinking, rinsing, and grooming), the physical therapist rated and recorded their levels of assistance as 1) independent, 2) modified Independent, 3) assistance required, or 4) dependent. Each participant performed the same activity (drinking, rinsing, or grooming) and sensor controls (voice, motion, or eye gaze) three times. For example, a subject used the voice control to activate fountain mode for drinking and repeated the procedure three times. A total of tests was 27 (3 sensors x 3 modes x 3 repetition).

Each activity was recorded "compete", "partially complete," or "unable to complete." The investigators also recorded the time it took for each subject to complete each activity.

After completing the feasibility testing procedures, each SCI subject completed the System Usability Scale (SUS) to help the investigators gain a better understanding of users' perceptions of the Access-H2OTM faucet's effectiveness, efficiency, and ease of use (Bangor et al., 2008). The detailed procedures for determining a SUS score are described in Bangor et al. (2008). Briefly, each subject rated 10 survey questions on a scale from "strongly agree" to "strongly disagree." Each response was converted to a numerical value from 1 to 5, and the SUS score was calculated using the formula (X+Y) x 2.5, where X = Sum of the points for all odd-numbered questions -5, and Y = 25 - Sum of the points for all even-numbered questions.