September 06, 2018 –
After a summer immersed in learning, conducting research and listening to talks presented by others, this year’s SURF-IoT Fellows finally got their moment in the spotlight. Twelve of the program’s 13 students shared the results of their 10-week research efforts by presenting final reports at the closing-day event for the Summer Undergraduate Research Fellowship in Internet of Things (SURF-IoT), held in the CALIT2 Auditorium on Thursday, August 30.
Sponsored by UC Irvine’s Undergraduate Research Opportunities Program and CALIT2, the SURF-IoT program offers students the opportunity to enhance their knowledge of telecommunications and information technology by conducting hands-on research side-by-side with faculty mentors.
Launched in 2005 as SURF-IT (Summer Undergraduate Research Fellowship in Information Technology), the program took on the IoT mantle three years ago to reflect the increasing importance of internet-of-things devices and applications.
A short summary of this year’s research presentations follows:
Cloudberry
Fellow: Yang Cao
Mentors: Qiushi Bai and Chen Li
The emergence of big data requires visualization systems that can help users understand the data and optimize queries. Cloudberry is a middleware system designed for fast data retrieval from huge datasets. Twittermap is a web application that visualizes a huge quantity of tweets by utilizing Cloudberry. Cao spent the summer improving Twittermap and developing new features for data analysis.
Distributed Mediator Placement for IoT Data Exchange Interoperability
Fellow: Andrew Chio
Mentors: Nalini Venkatasubramanian, Georgios Bouloukakis
As the number of IoT devices grows, so will the number of protocols employed by these devices to support direct internet connectivity. This project focuses on “mediators,” which convert and send messages from one protocol or device to another. Placement of these mediators is essential, especially in time-sensitive scenarios. Chio focused on analyzing where and how mediators should be placed — on the cloud, edge or fog – and on which physical device.
Improving Texera Interface for Powerful IoT Data Analytics
Fellows: Bolin Chen, Yuran Yan, Sibo Wang
Mentors: Chen Li, Avinash Kumar
Most data analysis applications are built for those with computer science backgrounds. Texera aims to provide a friendly, easy-to-use and powerful data analysis tool for non-computer scientists. The team reconstructed the code base of an old graphic user interface and implemented a new GUI framework, while polishing style and adding a user tutorial to make the workflow more user-friendly. The students also evaluated different workflow engines and software to fine the one most suitable for their implementation.
You Shall Not Pass: Applying Privacy Policies in Internet of Things
Fellow: Robin Cristobal
Mentors: Nalini Venkatasubramanian, Robert Yus (postdoc) Primal Pappachan (doctoral student)
Bluetooth, WiFi, HVAC sensors and other Internet-of Things devices have created smarter, more efficient buildings but they may have done it without regard for privacy. This project develops a system to allow those who enter IoT-equipped building to specify the context of their privacy and apply policies on incoming data. The system employs an enforcer engine, middleware that applies privacy policy actions as defined by a user before the data goes to the server.
Encouraging Healthy and Affordable Meal Planning
Fellow: Ling Jin
Mentor: Sergio Gago-Masague
This project involved the development of an application to help students create healthy, nourishing meals despite lack of money and/or time. The app combines recipes, ingredients and shopping information for meals that can be prepared in 15 minutes and cost less than $5 per person.
Identifying Personal Chronicles using Smartphone and Wearable Device Data Streams for Cybernetic Health
Fellow: Pooya Khosravi
Mentors: Ramesh Jain, Jordan Oh
This project’s goal is to help people connect to their smart devices to get insight into lifestyle, health and disease. Smartphones and wearable devices enable multimodal data streams from sensors to identify behavioral patterns and lifestyles but multiple data streams must be correlated. Personal Chronicles – personicles – synchronizes low-level data streams into atomic intervals of fixed size. Similar consecutive segments are clustered together into intervals to create daily segmentation. A key contribution of this research is fully automated sensor data collection and higher-level activity recognition to determine a person’s health state.
Single Cell Radio
Fellow Yongxi Li
Mentor: Peter Burke
A traditional method for detecting malfunctioning cells, organoids or organisms is to observe them under a microscope, which can be time-consuming and indistinguishable. This research proposes a chemiresponsive nanomaterial-integrated RFID- (Radio Frequency Identification) based technology aimed at identifying each organoid uniquely and tracking its physiological phenomenon effectively under an RFID reader, without using a microscope. The researcher approached this goal by integrating SWNT (single wall nanotube) material, which is chemically responsive, onto an RFID tag in order to render the tag detectable – or not – by the reader, depending on the chemical environment that the tag is in. The tag is inserted into a living organoid, and if the organoid malfunctions and releases certain chemicals, the tag becomes unreadable, enabling the observer to realize there is a malfunction in the cell.
OASIS: Assisting Low-Income Residents during Extreme Heat and Blackouts
Fellow: Barbara Martinez Neda
Mentors: Sergio Gago-Masague, Patricia Lim and Raquel Fallman
OASIS is a mobile application that aims to help people during severe weather conditions by providing them with the necessary tools to stay safe and hydrated. Libraries, community centers and senior centers often serve as cooling centers, and OASIS will have a map with the location and information of nearby cooling centers so users can utilize these services. Users will receive heat warnings and the app will also list information to prevent, detect and treat heat-related illnesses. With the proper knowledge, users will be able to reduce the number of emergency room visits as well as deaths caused by excessive heat exposure. With these features, the app aims to help low-income residents since they are often the most affected during extreme heat conditions. OASIS also will incorporate an alert system to remind users of the effects of leaving a child or a pet in an unattended vehicle.
Extreme Energy Use Optimization through Behavioral and Energy Usage Operational Cues
Fellow: Chen Mo, Mathias Wang
Mentors: G.P. Li, Michael Klopfer
Electric vehicles are becoming more and more popular; there will be 220 million of them on the road by 2030 according to estimates. The increased electricity demand could overwhelm the grid and lead to power failures as well as an increase in carbon dioxide emission. This project seeks to develop a large-scale smart EV charging control system that can optimize charging for millions of households using data management and prediction systems.
– Anna Lynn Spitzer
