ITRI will demonstrate several digital health technologies at CES 2020, including its WiFi Sleep Apnea Detection System and the companion robot for seniors PECOLA (Personal Companion for Older People Living Alone). At ITRI’s booth, visitors can experience in person how these innovations work. Below are the details of the two featured technologies.
The WiFi Sleep Apnea Detection System identifies the sleeper’s breathing status, using a low-power WiFi device to detect the respiratory rate. The user does not need to wear any device. When breathing is interrupted, an airbag cushion under the pillow slowly inflates. The slight height increase in the head position keeps the airway open and eases sleep apnea and reduces snoring. The height change of the inflatable cushion is about three-to-five centimeters, and the cycle of each rise or fall takes about one minute to complete, so the user will not notice changes in the pillow position. When the number of breathing stops decreases, the inflatable cushion is deflated back to its original state. ITRI’s solution allows the user to sleep comfortably while breathing is monitored.
The WiFi Sleep Apnea Detection System allows users to sleep comfortably while breathing is monitored.
The problem with existing contact-type wearable devices for detecting sleep apnea is that they require charging and are inconvenient and uncomfortable to wear, which results in low user acceptance. Moreover, contact-type devices have a fixed sensing range and often fail to measure reliably when patients roll over or move away from the sensing range during sleep. ITRI believes the WiFi Sleep Apnea Detection System with its non-contact breathing and heartbeat detection is the most intuitive and least intrusive for sleep monitoring. Use cases for the WiFi Sleep Apnea Detection System include homes and medical institutions. Health professionals can set up the device in the user's room to detect stopped breathing, respond accordingly to help reduce breathing pauses and snoring, and check daily sleep quality through smartphones.
To validate the accuracy of the system using the WiFi electromagnetic wave to measure breathing, ITRI worked with the Chang Gung Hospital Sleep Center in Taiwan to conduct human trials. ITRI found the average respiratory rate of the subjects in the Sleep Center measured by ITRI’s system was close to that measured by the ECG device. Moreover, the system can function normally without being affected by other working WiFi devices.
PECOLA (Personal Companion for Older People Living Alone) is a companion robot that applies ambient intelligence technology to seamlessly measure the day-to-day life quality of seniors living alone. It provides such information about the seniors as “do they eat healthily?”, “do they sleep well?”, “are they in any bad mood?”, and “do they exercise enough?” to enable the seniors’ friends and families to engage more deeply with them.
When detecting a fall or abnormal behavior, PECOLA will reach either a caregiver or emergency contact.
PECOLA is an autonomous mobile robot shaped like a moving piece of furniture which consists of a camera, a WiFi access point, an auto phone dialer, and a microphone array. The camera is equipped with intelligent video analysis capabilities such as face detection, expression recognition, diet analysis, physical activity assessment, and fall detection. If a fall or abnormal behavior is detected, the system reaches either a caregiver or emergency contact. The WiFi AP offers a transparent way to accurately measure breathing and heart rate without requiring wearable sensors. An array of small microphones collects multidirectional audio data to detect keywords for emotion analysis and to identify sources of specific sounds. For autonomous navigation, the mobile robot is equipped with multiple obstacle detection sensors, including a camera and 2D LiDAR, for smooth maneuvering in home environments.
With ambient intelligence at its core, PECOLA supports multiple advanced AI-based functionalities. Sleep Analysis: WiFi-based breathing and heart rate detection technology; Diet Analysis: Video-based food portion change, fresh versus leftover, specific dish name (currently Chinese food only), and nutrient-content detection; Activity Analysis: Four human poses (standing, sitting, bending and lying), video-based fall detection technology, interval walking training program for blood sugar control; Emotional Analysis: Video- and voice-based detection of emotions: happy, angry, sad, surprise, fear, disgust, contempt, and neutral emotions.
Although many devices and services have emerged recently to help seniors age healthily, PECOLA is uniquely and specifically designed to cater to two of the most important needs of seniors’ friends, families and caregivers. First, PECOLA provides peace-of-mind for them when they can’t be physically present. Second, PECOLA provides them appropriate topics of conversation for meaningful interaction and social engagement. Through ambient intelligence technology, PECOLA measures various quality of life indicators of seniors living alone and detects emergency situations without the need to wear any sensors or install sensors around their home.