1st Place Winner: Enhancing Healthcare Access through Remote Infant Screening
University: United States International University – Africa
Team Members: Khushi Gupta, Dharmik Karania, Jeet Gohil, Abdihamid Ali
Faculty Advisor: Dr. Leah Mutanu
The growing interest in automating many aspects of our daily lives has seen an increase in the use of emerging technologies such as Internet of Things (IoT) to address numerous societal challenges. A key challenge in many developing countries is health access, especially for people in rural or marginalized communities. In this project we looked at how we can leverage on the benefits of IoT technology to reduce the high infant mortality rate brought about by poor health access. Babies are required to go for regular check-ups from birth until the age of 24 months to detect growth and general health issues for timely intervention. However, many parents are unable to do this due to the cost of this check-up, and long distances to the health centres. Health centres are also very few and constrained with the lack of adequate personnel and resources required to conduct this screening.
The innovation involved the design of a system used for remote monitoring of infants’ growth and health parameters. In most rural areas in Kenya, access to basic child health services often requires a long journey to the nearest health facility. The cost of the journey, in addition to a small fee charged at the health facility, is out of reach for most rural families. The net result is that many families in remote rural areas do not visit health facilities until it is too late. Thus, many children in rural areas continue to suffer from preventable health related problems that could easily be addressed. Consequently, under-five mortality remains unacceptably high at 46 deaths for every 1000 births according to UNICEF. The developed solution provides community health workers, or care givers, with a portable, safe, and accessible Internet of Things (IoT) device that can be used to monitor growth parameters of children such as height and weight and health parameters such as skin condition and temperature. This data is stored in the cloud and used to generate growth charts reports or send alerts to health facilities on children who require immediate attention for timely intervention.
2nd Place Winner: Driver Distraction System: A Deep Learning Approach
University: Mansoura University
Team Members: Abdulrahman AbouOuf, Omar AlEzaby, Omar AlSaqa, Mohammad Nasser
Faculty Advisor: Dr. John Fayez
This project aims to reduce number of car crashes by using AI systems that help the driver not to get distracted. Our solution is to make everything inside the cabin accessible by voice (distraction elimination). Also we believe that our project would make it possible to ascertain quickly drivers’ actions and their sights and to continually build use cases so we can understand when something may interfere with the driving process and create new ways to alert the driver (distraction detection). Our project is composed of 5 AI/ML components: 1- Driver actions classification: which is done by a camera takes images of the driver and classify his action whether he’s distracted or not. 2- Head pose estimation: which is done by another camera takes images of the driver’s head and provide its angles. 3- Speech to text Engine: to generate text transcriptions from the driver voice to prepare it for the text classification model. 4- Text classification: which takes the text transcription and classify it to which command the driver wants to do. 5- Trigger word detection: to run the whole system by voice without getting distracted. After a complete compression study of how we compress all these models together on a chip board we combined all these tasks together and we came up with a complete new multi-model system for driver distraction elimination and detection. What’s interesting about this project is that while developing, testing and enhancing we weren’t only concentrating on the technicalities, but we kept in mind the end goal of the project which is to create a business product that works offline to provide safety to the driver.
3rd Place Winner: Analysis of students’ activity on e-learning course based on OpenEdu platform logs.
University: Peter the Great St.Petersburg Polytechnic University
Team Members: Nikita Barsukov, Ivan Sysoev, Vladimir Ermoshin
Faculty Advisor: Dr.Igor Nikiforov
The project revolves around analysing the logs of platforms based on an open-source online education platform “OpenEdx”. One of such educational projects is “Open Education”. Our software solution solves a large number of analytical tasks of student activity on the courses such as cheating on the tests, the ability to understand the popularity of video lectures, the ability to see the user’s active time on the course, and others. The analytic platform of “OpenEdx” which is provided out of the box, is limited in functionality and has little flexibility. The undoubted advantage of analysing the logs is the direct work with platform logs. It allows us to build flexible analytics at the request of the teacher. Course logs can be obtained directly from the platform using the provided API, or they can be downloaded directly from the archive. Usually, the volume of the logs is from 3 to 20 gigabytes. After that, we upload log files in the database and work directly with them
Learn more about Envision The Future here.