Mentor Areas
My research interests include cyber-physical systems, real-time and embedded systems, safe autonomy, runtime assurance and verification, internet of medical things, and connected health. The theme of my research activities has been to assure and improve the safety, security, and timeliness of life-critical embedded systems.
Description:
1. Time-sensitive Networking (TSN) Test-bed setup and experiments.
Time-sensitive Networking (TSN) is a set of standards under IEEE 802.1 working group, aimed to provide a transmission of low latency and high availability in bridges or bridged networks (IEEE 802.1Q).
The goal of our project is targeted to generate TSN schedule that can provide guaranteed network QoS in the transmission. We are building a test-bed for Time-sensitive networking (TSN) to perform some experiments of a new TSN scheduling algorithm. We are looking for someone to help set up the TSN Test-bed. The work includes installation of embedded Linux, IP network configuration, time synchronization (PTP), DPDK programming, and TSN schedule configuration. It may also need Linux kernel programming/debugging.
2. Medical device dashboard.
We are building an integrated medical device dashboard that can help to maintain the medical devices from various vendors and to troubleshoot them. The system is composed of front-end and back-end. The front-end is a website written in ReactJS. The back-end is providing data access from the database (MongoDB) via REST-API or WebSocket, and the back-end application is written in Java (Spring-Boot).
We are looking for someone to extend the back-end REST-API/WebSocket according to changed/added data in the database or to upgrade the website such as implementing additional web pages with new data or improving the existing pages with a better graphical representation.
3. DNN-based perception systems are difficult to assure because they map pixels directly to high-level outcomes, such as low-dimensional numbers. Existing analysis tools have limited scalability for such systems, and the DNNs for perception are typically not trained to pass this verification. This project investigates how to design and assure a sonar-based perception system in an autonomous underwater vehicle. This goal of this perception is to estimate the range from the vehicle to an underwater pipe (which is subsequently used for navigation and control). We design and analyze this perception system based on a ROS/Gazebo simulation in three steps: (a) generation of realistic sonar noise and collecting a representative dataset, (b) training a neural network to detect pipeline presence in a small sub-image of sonar scans, and (c) estimating the range to the pipe from raw sonar data by "sliding" the detection network over a full sonar scan.
Preferred Qualifications
1. C language, Basic knowledge of computer network (TCP/IP), socket programming, Linux kernel programming (optional)
2. Front-end: Web-programming (HTML, NodeJS, and ReactJS);
Back-end: Java programming (Spring-Boot), and MongoDB
3. Machine Learning
Details:
Preferred Student Year
Second-Year, Junior, Senior
Volunteer
Yes
Yes indicates that faculty are open to volunteers.Paid
Yes
Yes indicates that faculty are open to paying students they engage in their research, regardless of their work-study eligibility.Work Study
Yes
Yes indicates that faculty are open to hiring work-study-eligible students.