As part of the MSTI curriculum, students work in small teams with GIX Consortium members, partners, and UW faculty members to create hardware and software prototypes that solve real-world problems. This year, ten teams worked on projects using emerging technologies in areas including global population health, wildland firefighting, and memory therapy.
Team AMS
Team: Yingru Feng, Cody Gagnon, Jerry Liang, Yujie Zhu, Yuzhou Zhuang
All commercial jets use an Auxillary Power Unit or APU to provide power and preconditioned air to the plane when parked at the gate. APUs are necessary for the function of jets, but they aren’t needed when greener sources of energy are available. In partnership with Microsoft’s AI for Earth initiative, the team developed a system which automatically monitors APU usage in real-time, providing operational insights into a previously inaccessible data stream, and enabling enforcement of upcoming environmental regulations.
Team Cocobot
Huichen Li, Yun Liu, Lu Wang
Cocobot is a virtual therapy chatbot that helps family caregivers of people with chronic conditions access self-care tools more conveniently and flexibly. We designed mechanisms for Cocobot to learn users’ behaviors over time by collecting users’ feedback for each session, track users’ practice history, and monitor users’ emotional changes. With customized experiences and flexible interactions, we encourage users to stay engaged and form long-term selfcare habits.
Team Direct
Frank Zhou, Ketki Hardas, Jialin Zhao, Lu Wei, Tommy Ho
Diabetic Retinopathy (DR) is the leading cause of blindness not only in the USA but also in developing countries. 370 million people worldwide suffer from diabetic retinopathy, glaucoma, or macular degeneration. The proposed solution is a portable low-cost fundus camera with AI-based software + supporting web app to facilitate data collection.
Team Fireworks
Paulo Goncalves, CJ Ngeh, Jiali Zhang, Gulnara Sarymbekova, He Feng
Heat-related illness is a severe problem that wildland firefighters face during their deployment. Since 2000, 313 wildland firefighters have lost their lives on the field, and 54% of them were caused by overexertion under hot and challenging conditions. Our propsed solution involves using heart rate, temperature, and movement data to alert firefighters and their supervisors about potentially hazardous heatrelated illnesses. We have developed a smart helmet with integrated sensors to monitor the health data, a corresponding FireWorks App to show data visualization to the wildland firefighters and supervisors, and an alert system on the helmet to nudge the firefighters and supervisors when it detects abnormal data.
Team HealthEar
Dominick Mendoza, Lyuting Wu, Robin Yang, Zhan Shi
80% of children will have had at least one episode of an ear infection by the age of 2. Guardians will lose sleep, miss work to schedule appointments, and have to watch their child in pain during these ear infections. There is also a large financial impact, where ear infections are
responsible for 24 million office visits per year costing $2 – 5 billion dollars annually. Using commercially available wireless earbud headphones to aid in the diagnosis of ear infections in children by sensing changes in the acoustic properties of the eardrum.
Team Memory Lane
Amal M. Abualrahi, Joey Wang, Yi Zheng
People living with dementia become less social and self-isolate as they recognize their memory loss, experience time confusion, and anxiety due to feeling they are in the wrong place. These dementia-related symptoms cause challenges for caregivers and family members. Memory Lane is a platform that allows users to navigate and annotate digital content using tangible interactions.
MemoryLane provides:
- High accessibility for people living with mild to moderate dementia.
- Easy and simple setup process; No pairing is required.
- Automated customization using Machine Learning.
- Simple caregivers web portal to access users’ profiles and data.
Team SCIO
Amy Chen, Kelsey Guo, Ken Christofferson, Shi Ni, Steven Guh
More than a billion rapid diagnostic tests (RDTs) are administered each year to screen for some of the world’s most deadly illnesses. However, there are significant opportunities to improve RDT result accuracy and provide that data to decision-makers more quickly. Microsoft and PATH partnered to improve RDT outcomes by creating a universal mobile phone based RDT reader. The GIX SCIO team kicked off work by creating an image collection application based on laboratory technician workflow and a machine learning solution that can read RDT results in a limited number of RDT models.
Team Spyro.AI
Arpan Agarwal, Jay Chakalasiya, Chuck Scott
Asthma is a global health concern and one of the most prevalent diseases in the world. It impacts 1 in 7 people and can dramatically restrict activities of daily life, up to debilitation and death. Asthma is a highly individualized condition, with regards to triggers and treatments. The correlation of respiratory health measurements with physical and environmental factors can play a critical role in the personalization of care and treatment. Our solution brings new forms of longitudinal data to physicians to inform their decisions about how they treat patients with asthma. In the long-term, this could be used to discover new insights about the disease and improve patient health over their lifetime.
Team Dronected
Issac Boger, Justin Ho, Ke Wang, Chun-an Ku, Wenbo Zhong
Drones are an increasingly vital tool for search and rescue operations, but they still have serious limitiations. Pilots must split their concentration between flying the drone and scanning the video feed, or they must physically transfer higher quality footage to a computer, wasting precious hours when lives are on the line. By leveraging the power of 5g networks, high-bandwidth data feeds allow a new cooperative workflow where the pilot can focus on flying and S&R experts can assist in navigation and spotting from anywhere in the world.
Team T-Mobile IoT
Alvaro Carmona, Chen Ma, Vivian Wang, Yida Zhao, Wenyuan Zhang
IoT (Internet of Things) are at a blooming stage, with a growing amount of IoT devices, the demand for testing and certifying these devices skyrocketed as well. Current certification process takes 6-9 months for a device to be fully certified, including 4 months of pre-test. This is affecting business results for T-Mobile and its partners. We created a self-service certification toolkit to help developers pre-test their IoT devices before sending them into labs. This solution would save both parties time on device transferring also ensure certain quality assurance. Our solution allows developers to create and execute test cases based on AT commands this will then verify whether the devices pass certain requirements to operate on T-Mobile’s network.