Caring for my grandfather, who is afflicted with Alzheimer’s disease, has caused our family significant stress, particularly when he wanders out of bed at night and suffers accidents. My grandfather is one of the 5.2 million Alzheimer’s patients in the U.S., 65% of whom wander. To protect their safety and alleviate the burdens on their caregivers, I invented a low-cost wearable sensor technology for real-time, reliable detection of patients’ wanderings. Once the patient steps onto the floor, a sensor worn on the foot will immediately detect the pressure caused by body weight and wirelessly trigger an audible alert in a caregiver’s Smartphone. I developed three enabling technologies: an ultra-thin film sensor that is comfortable to wear, a coin-sized wireless circuit enabled by cutting-edge Bluetooth Low Energy, and apps that transform Smartphones into caregivers’ monitors. Integrating them, I created two prototypes: a sensor sock and a sensor assembly that can be conveniently adhered to a foot. A six-month trial on my grandfather validated my hypothesis: the systems detected 100% of the 437 known cases of his wandering and issued alerts within one second of his stepping out of bed. No false alarm was issued. I am in the process of testing my technology in nursing homes and investigating patterns from sensor data. In addition to solving the originally intended problem, using the sensor to monitor a larger population of Alzheimer’s patients could lead to a fundamental understanding of the causes of wandering and thus ways to mitigate or prevent it.
I love exploring, whether it be the Grand Canyon or a tiny integrated circuit. As an Eagle Scout, I enjoy outdoor activities. As a lover of technology, I never get tired of dreaming in my small bedroom about the next big innovation. Growing up in a three-generation family burdened with caring for my grandfather, who suffers from Alzheimer’s, I have always been concerned about the wellbeing of my grandparents. At age 6, I invented a Smart Bathroom that sends a buzz alert to the wristwatch of children when their elderly parent falls down in the bathroom. A year later, I created a Smart Medicine Box that emits a sound and flashing light to remind patients to take the right medicine at the right time. Both inventions won first place in local science/invention competitions and further sparked my interest in technology. I admire all the great innovators who have changed the world, but my direct inspiration comes from my parents, both civil engineering professors committed to making structures healthier and safer. However, I am more interested in the health of the human brain. I aspire to bridge engineering and neuroscience by inventing technology to explore the mysteries of the brain and find a cure to Alzheimer’s. Winning the Google Science Fair will shine a spotlight on my technology, making it available to the millions in need. Analyzing data from this large population of patients will enable me to investigate the causes of wandering, bringing me one leap closer to my dream.
Question / Proposal
On a chilly August dawn in 2005, my grandfather, dressed in his pajamas, showed up at our front door with a policeman. We were shocked to learn that Grandfather had been wandering aimlessly on a freeway two miles away from our home, while we were sleeping. Subsequently, he was diagnosed with Alzheimer’s disease and the dynamic of my three-generation family changed forever.
We learned that my grandfather is just one of the millions of Americans who wander because of Alzheimer’s, dementia, and other mental diseases. According to the Alzheimer’s Association, 5.2 million Americans suffer from Alzheimer’s. Sixty-five percent of these patients wander, which, especially when happening at night, causes significant stress to their 15.4 million caregivers (whose service was valued at more than $216 billion in 2012 alone).
I grew increasingly concerned about my aunt, Grandfather's primary caregiver, because she had to wake up frequently every night to check on him. Despite her efforts, she was seldom able to catch his wandering, and thus he suffered many accidents. We searched extensively for a device that could detect bed-wandering, but nothing was helpful.
The lack of an effective solution motivated me to invent a wearable device to detect patients’ wanderings out of bed and alert their caregivers. My hypothesis is that the system would need to reliably detect and issue an alert within three seconds of the patient’s stepping out of bed, without any false alarm. The system would need to be affordable, easy to operate, and powered by coin battery.
Existing technologies did not provide a solution to my problem. For instance, widely used GPS tracking devices do not work indoors. More recent real-time locating systems based on radio-frequency (RF) ID tags are neither accurate nor quick enough to detect patients’ leaving bed. Initially, I considered attaching pressure sensors to the legs of a bed, but the system would be expensive and could trigger a false alarm from my grandmother, who was sleeping on the same bed as my grandfather.
One night, I was looking after my grandfather and saw him step out of bed. The moment his foot landed on the floor, a light bulb flashed in my head. Why don’t I put a pressure sensor on the heel of his foot? The moment he steps onto the floor, the sensor would detect the pressure caused by his body weight, and the signal could wirelessly trigger an audible alert in my aunt’s Smartphone. This would enable my aunt to sleep much better at night without worrying about Grandfather.
I immediately put down my thoughts in a sketch (Fig. 1) and started research. First, I had to find a pressure sensor that was thin and flexible enough not to affect the comfort of walking. An extensive literature search led me to a film sensor printed with pressure-sensitive ink. When pressure is applied, the ink becomes more electrically conductive. The pressure can thus be calculated by measuring the conductivity. I also thought about creating a sensor out of conductive rubber, but found it difficult to make such a sensor thin enough after testing material samples.
My second challenge was designing a wearable circuit to acquire signal from the pressure sensor and wirelessly trigger an alert in a remote Smartphone when the pressure exceeds a certain threshold. The circuit needed to be compact, lightweight and human-friendly. However, the energy-consuming wireless transmission requires large, heavy batteries. A search of state-of-the-art wireless systems led me to Bluetooth Low Energy (BLE), a rapidly-evolving, revolutionary wireless technology that has recently enabled the creation of many novel wearable systems. Implementing BLE required a significant amount of expertise, which I had to acquire through extensive studying and testing of BLE modules. After months of research, I was able to design a circuit that would integrate such a module with my film sensor and be powered by a coin battery.
In addition, I needed to design apps to transform a Smartphone into a caretaker’s monitor. I took online courses and tutorials to learn coding for the most widely-used Android and iOS phones. With the help of a mentor, I deepened my knowledge about the core Bluetooth framework, which enabled me to establish communications between the BLE circuit and Smartphone while minimizing energy consumption.
This research convinced me of the feasibility of my proposed wearable sensor system, thanks to the recent advances in Smartphones and BLE wireless technologies. I made a one-year project plan with five major tasks to realize my proposed sensor system (Table 1).
Method / Testing and Redesign
Task 1: Film Sensor
Step I - Conductive ink and sensor prototyping: Applying what I read from scientific papers, I devised a composition formula for the sensor ink: a mixture of electrically conductive particles and resin. I acquired and mixed the materials and printed sensor samples using a silk-screen printer.
Step II - Sensor peeling test: I conducted a peeling test to evaluate the bonding strength of the ink to the film, which I found to largely depend on the type of resin. In addition to selecting a resin, I also applied physical and chemical treatments to the film surface to further enhance the bonding.
Step III - Sensor calibration: I carried out calibration tests by placing the film sensor next to a reference piezoelectric force sensor and applying a mechanical force at different values (Fig. 2). To ensure consistency, I repeated each test three times. Based on the results and recommendations from literature, I further improved the ink formula until an ideal linear relation between the pressure and the conductance was achieved.
Task 2: Wearable Wireless Circuit
Step I – BLE module evaluation: The lack of performance comparison information prompted my evaluation tests of available BLE modules (each containing a microcontroller and an antenna). I carried out 16 cases of measurement to evaluate the wireless signal strength of three newest BLE modules, based on which I selected the most cost-effective BLE module.
Step II – Integration of BLE with sensor: From a mentor, I learned to design and make a printed circuit board to integrate the BLE module with my sensor and a coin battery. I tested a number of designs with different antenna positions/orientations to maximize the wireless signal strength. I was able to shrink the entire circuit to a board the size of a quarter coin (Fig. 3).
Task 3: Smartphone Apps
Step I – Triggering Algorithm: I made a flowchart for the apps (Fig. 4) that would command the circuit to acquire sensor signals at a predetermined frequency (e.g., 1Hz), compare it with a threshold, and send a wireless signal to trigger an audible alert in the Smartphone, once the threshold is exceeded. In order to investigate the reliability of the triggering and determine the threshold value, I tested the sensor on family members of different weights stepping on floors with various surfaces (Fig. 5).
Step II – Coding: I coded and tested apps for transforming Android and iOS phones into caretakers’ monitors.
Task 4: Prototypes
Integrating the above technologies, I produced a sensor sock embedded with the film sensor and the circuit board (Fig. 6). Realizing that some patients may not wear a sock during sleep, I developed a second prototype: a sensor assembly that can be conveniently adhered to a foot or a sleeper (Fig. 7).
Task 5: Evaluation Test
Both prototypes were tested on my grandfather every night for six months to evaluate their reliability, response speed, ease of use, and durability.
Enabling Technology #1 - Film Sensor: I developed 0.25mm-thin flexible film sensors to accurately measure pressure. Figure 8 shows the calibration test results on two of the film sensors made in this project. The measured resistance and conductance, averaged from three measurements, are plotted with respect to the applied force (measured by the reference sensor). As expected, the conductance of the sensors increases as the force increases. Sensor #2, printed with the improved ink formula, exhibits a better linear relation between the conductance and the force than Sensor #1, validating the ink design. Using this calibrated relationship, the applied force (or pressure when divided by the sensor area) can be calculated by measuring the conductance (or resistance) of the sensor.
Enabling Technology #2 - Wearable Wireless Circuit: Using a Smartphone as a receiver, I tested three of the newest BLE modules by placing them 50 ft apart, each in four different orientations, resulting in 48 cases as shown in Fig. 9. To eliminate environmental effects on the signal strength, I measured at the same time in the same room. Figure 10 plots the received RF signal strength time histories for the 48 cases. BLE112 showed the strongest signal, followed by BLE113. By quantitatively evaluating all the factors including power consumption, size, and software development environment, as shown in Table 2, I selected the most cost-effective module BLE113 to use in my sensor circuit. I achieved a circuit the size of a quarter coin (Fig. 3). It is powered by a coin battery, which would last for at least one year assuming daily usage, according to my calculation.
Enabling Technology #3 - Smartphone Apps: I developed apps for the widely used Android and iOS phones to enable them to serve as caregivers’ monitors. In order to prevent a false alarm, the triggering algorithm (Fig. 4) was further improved to compare the sensor output signal with the threshold at three consecutive time instances, rather than only at one. Many techniques were implemented into the software to minimize the wireless power consumption.
Prototypes and Evaluation Tests: Integrating the enabling technologies, I made two prototypes of the wearable sensor system. During the six-month trial on my grandfather, the prototypes successfully detected 100% of the 437 known cases of his wandering out of bed at night. The alerts were issued within one second after he left bed. The app also recorded the time of each alert. The data, summarized in Table 3, show patterns: each and every night, he wandered out of bed at least twice; for most (71%) of the nights he left bed only twice (around 1:00am and 4:00am); and occasionally (6% of the nights) he wandered as many as four times per night.
There were no false alarms issued when he did not leave the bed. The prototypes were found to be easy to use and durable, surviving the 6-month nightly usage. During the trial, my grandfather suffered no accidents at night, and my aunt was able to sleep much better.
Conclusion / Report
I have succeeded in developing my proposed technology into two working prototypes and validating my hypothesis of using a wearable sensor to reliably detect a patient’s wandering out of bed and wirelessly trigger an audible alert in his caregiver’s Smartphone in real time. To make the system wearable, I created a 0.25mm-thin (one tenth the thickness of the human skin) film sensor and a coin-sized ultra-low-energy wireless circuit. I also made apps to enable Android and iOS Smartphones to serve as caregivers’ monitors.
I carried out extensive tests. For the film sensor printed with pressure-sensitive electrically-conductive ink, I found a composition formula for the ink that resulted in an ideal linear relation between the applied pressure and the measured conductance. For the wireless circuit, I selected the most cost-effective BLE module through extensive evaluation tests and integrated it with my film sensor. For the apps, I developed, tested, and improved an alert triggering algorithm. Integrating these three enabling technologies, I produced two prototype systems: a sock embedded with the sensor and a sensor assembly that can be adhered to a foot, sock or shoe. During the six-month trial, the prototypes successfully detected 100% of the 437 known cases in which my grandfather wandered out of bed at night. The alerts were triggered within one second after he stepped onto the floor. There were no false alarms triggered. During the trial, he suffered no accident and his caregiver reported that she slept much better when using the sensor.
This project has not only validated my hypothesis, but also opened up new avenues of research. Using the sensor data from my grandfather, I am studying the correlation between the frequency of his nightly wandering and his daily diet/activities. Such information could be useful for developing a predictive and preventive care. This approach could have wider applications. For instance, using this sensor to monitor a larger population of Alzheimer’s patients could lead to a fundamental understanding of the causes of wandering and thus ways to mitigate or prevent it. This has further motivated me to validate my technology on a large population of patients. I am in the process of having hundred of units manufactured for donation to nursing homes, where I have provided community service with my fellow Boy Scouts to Alzheimer’s patients.
This project has given me an incredible life experience - the excitement of creating something new, the pain of getting nowhere, the joy of overcoming obstacles, the pride of accomplishment (including my U.S patent and journal publication from this project), and the peace of mind from knowing that my grandfather is safe and my aunt can sleep better at night. I will never forget how deeply moved my entire family was when they first witnessed my sensor detecting Grandfather’s wandering. At that moment, I was struck by the power of technology to change lives. I am now even more motivated to pursue my passion for technological innovations that solve health care problems facing our increasingly aging society.
Bibliography, References and Acknowledgements
Abellan Van Kan, G., Rolland, Y., Andrieu, S., Bauer, J., Beauchet, O., Bonnefoy, M., … Vellas, B. (2009). Gait speed at usual pace as a predictor of adverse outcomes in community-dwelling older people, an International Academy on Nutrition and Aging (IANA) Task Force. The Journal of Nutrition, Health & Aging. 13(10), 881-889. Retrieved from http://link.springer.com/article/10.1007%2Fs12603-009-0246-z
Alzheimer’s Association (2013), 2013 Alzheimer’s disease facts and figures. Retrieved from http://www.alz.org/downloads/facts_figures_2013.pdf
Apple (2013). Programming with Objective-C, Mac Developer Library. Retrieved from https://developer.apple.com/library/ios/documentation/cocoa/conceptual/ProgrammingWithObjectiveC/Introduction/Introduction.html
Ashruf, C.M.A. (2002). Thin flexible pressure sensors, Sensor Review, 22 (4), 322 – 327. doi: 10.1108/02602280210444636
Bluegiga (2012). Bluetooth smart – Getting started. version 1.4. Retrieved from http://teleorigin.com/download/Bluetooth/Low%20Energy/BLE_getting_started_v1.4.pdf
Boulos, M. N. K. and Berry, G. (2012). Real-time locating systems (RTLS) in healthcare: a condensed primer. International Journal of Health Geographics, 11 (25), 1-8. Retrieved from http://www.ij-healthgeographics.com/content/pdf/1476-072X-11-25.pdf
Ching, C. (2013). XCode tutorial for beginners, Retrieved from http://www.youtube.com/watch?v=LTsAZSGlWgg
Faucounau, V., Riguet, M., Orvoen, G, Lacombe, A., Rialle, V., Extra, J., and Rigaud, A.S. (2009). Electronic tracking system and wandering in Alzheimer’s disease: a case study. Annals of Physical and Rehabilitation Medicine, 52(7-8), 579-587.
Rangel, J., del-Real, A., and Castano, V. (2008). Smart Conductive Inks. Chemistry and Technology, 2(4), 305-308.
Hardy, S.E., Perera, S., Roumani, Y.F., Chandler, J.M., Studenski, & Stephanie, A. (2007). Improvement in usual gait speed predicts better survival in older adults. Journal of the American Geriatrics Society. 55(11), 1727-1734. Retrieved from http://www.interactivemetronome.com/imw/IMPublic/Research/Temporal%20Processing/Gait/Research_GAIT%20MATE_improvements%20in%20usual%20gait%20speed%20predicts%20better%20survival.pdf
Heydon, R. (2009), Bluetooth low energy: The developer’s handbook. Prentice Hall, Upper Saddle River, New Jersey.
Hirasawa, M., H. Okada, M., Shimojo, M. (2008). The development of the plantar pressure sensor shoes for gait analysis, Journal of Robotics and Mechatronics, 20 (2), 289-295. Retrieved from http://www.fujipress.jp/finder/preview_download.php?pdf_filename=PRE_ROBOT002000020012.pdf&frompage=abst_page&pid=1420&lang%3B=English.
Kelly, K. (2013). Who wanders – Plain talk about Alzheimer’s. Alzheimer’s America, Retrieved from http://www.alzheimersamerica.com/wandering.html
Kochan, S. (2012), Programming in Objective-C, 5th Edition, Developer’s Library, Informit.com
Luo, Z-P, Berglund, L, & An, K-N. (1998). Validation of a F-scan pressure sensor system: a technical note. Journal of Rehabilitation Research and Development, 35 (2), 186-191. Retrieved from http://www.rehab.research.va.gov/jour/98/35/2/pdf/luo.pdf
Razak, A.H.A., Zayegh, A., Begg, R., & Wahab, Y. (2012). Foot plantar pressure measurement system: a review, Sensors, 2012 (12), 9884-9912. Retrieved from www.mdpi.com/1424-8220/12/7/9884
Sabban, A. (2013). Comprehensive study of printed antennas on human body for medical applications. International Journal of Advance in Medical Science, 1 (1), 1-10. Retrieved from www.seipub.org/ams/Download.aspx?ID=2283.
Wimo, A. and Prince, M. (2010), World Alzheimer report 2010: The global economic impact of dementia. Alzheimer’s Disease International, London, U.K. Retrieved from http://www.alz.co.uk/research/files/WorldAlzheimerReport2010.pdf
Dr. Alan Kaganov, a successful researcher-turned entrepreneur with deep expertise in biomedical engineering and healthcare technology, has kindly mentored me on sensor design, project development, and business development. He encouraged me to reach out to more Alzheimer’s patients to study their needs and test my technology. He has provided me with significant recourses such as an access to a top manufacturer for future mass production, as well his thoughts on research directions, patent application and journal publication.
I could not have completed this project without the invaluable advice and help of Mr. Masato Mizuta. He suggested reading materials that gave me crucial knowledge about microcontroller/BLE circuitry and XCode. His thoughtful suggestions and constructive criticisms of my designs played a critical role in my project. In addition, Mr. Mizuta helped me with the layout design and fabrication of the printed circuit board integrating the BLE module and programming of the microcontroller, all of which required high-level technical skills only be acquired from years of hands-on experience. From Masato, I not only learned how to create an engineering system, but also how to find resources.
A member of the National Academy of Engineering, Dr. Nick Alexopoulos is one of the world’s leading scholars in electromagnetics. I benefited significantly from his advice on my energy-efficient wireless circuit design, in particular improving the wearable antenna. More importantly, he taught me how to approach and carry out research in general. From Dr. Alexopoulos, I have learned how to think outside of the box.
Dr. Shigeyuki Yoneyama, a former intellectual property executive at Sony Corporation, helped me to submit a patent application in Japan. During the process, he made a number of constructive suggestions that broadened the technical scope of my invention.
Antonio Camacho, Scoutmaster of Boy Scout Troop 729, introduced me to several nursing homes in New York City, where I am engaging my troop to prepare trials of my technology on their Alzheimer’s patients.
Last but not least, my family’s influence, encouragement and unconditional support have made this work possible. Since I was very young, my parents have always encouraged me to expand my curiosity by exploring the world on my own. As my role models, they have sparked my interest in discovering problems in society and finding creative solutions. My aunt and grandmother have helped me to test my prototypes on my grandfather and have provided me with many practical suggestions for design improvement.
My work that required specialized equipment and tools, including silk-screen printing of the film sensors, sensor calibration tests, and fabrication of the printed circuit board, was carried out in a lab in the School of Engineering, Columbia University. The 6-month trial was carried out in the bedrooms of my grandfather and my aunt. All other work was done in my own bedroom, including online literature search, system design, sensor and circuit designs, circuit testing, monitoring data analysis, and app development.