Ashwin Srinath, Clemson University, June 15, 2018

Researchers in Civil Engineering are interested in exploring the computing infrastructure that would be critical to support a data-rich future for connected vehicle transportation systems. Clemson ACI-REFs assisted in the design and implementation of such an experimental infrastructure. We also facilitated and trained Civil Engineering students in using large-scale computing resources, on-site and national, to deploy this infrastructure for research activities.

The Objective

Since 2016, a Civil Engineering research group led by Dr. Mashrur Chowdhury has explored the challenge of moving and integrating massive amounts of data coming from various sources (such as roadside and roadway equipment, social media, online weather resources, etc.,) to a large-scale computing infrastructure. This is a new research direction aimed at the upcoming deluge of data that results from a future society where everything is connected. To date, most Civil Engineering research in the area of connected transportation systems has focused primarily on the transportation applications themselves, such as… . Less research has addressed the backbone of cyberinfrastructure that is critical to these transportation applications.

One challenge for this area of research is to evaluate various cyberinfrastructure designs quickly and at relatively large scale for validation purposes. This is important because… A shared-memory solution is not appropriate, as the individual computing entities need to reside on isolated physical devices. Another challenge is to ensure the maintainability and reproducibility of these experiments for future work. Dr. Chowdhury’s group thought working with ACI-REFs would help them address these challenges…?

The Solution

Clemson University’s ACI-REFs Linh Ngo and Ashwin Srinath worked with graduate students and post-doctoral researchers from Dr. Chowdhury’s research group to identify potential solutions to these challenges. The scope of assistance that Dr. Chowdhury’s students received include training and consultation services (including the Palmetto cluster?). For training, the researchers learned Python programming, Linux, and how to use Clemson’s research computing resources. For research consultation, the facilitators worked with the researchers to help them understand a range of topics, including the temporal and spatial requirements of connected vehicles transportation, and back-end computing infrastructure performance metrics. The facilitators also introduced the research group to CloudLab, a cloud-based public computing resource that enables the deployment of experimental computing infrastructures. This allowed them…

Using CloudLab and Clemson University’s Palmetto, Dr. Chowdhury’s research group was able to design and deploy a large-scale experimental environment including more than a hundred computing instances running on isolated hardware instances located on different host institutions.

The Result

This environment allows… Through this environment, a prototype data delivery infrastructure was tested and results published in a peer-review journal and presented at a professional conference.

Notable Publications and Presentations Resulting from this Work

Du, Yuheng, Mashrur Chowdhury, Mizanur Rahman, Kakan Dey, Amy Apon, Andre Luckow, and Linh Bao Ngo. “A Distributed Message Delivery Infrastructure for Connected Vehicle Technology Applications.” IEEE Transactions on Intelligent Transportation Systems (2017).
Rahman, Mizanur, Yuheng Du, Linh Bao Ngo, Kakan Dey, Amy Apon, and Mashrur Chowdhury. “An Innovative Way to Manage Data for Connected Vehicle Application.” Poster Presentation at Transportation Research Board 95th Annual Meeting (2016).

Collaborators and Resources

• Mashrur Chowdhury, Professor, Civil Engineering, Clemson University
• Amy Apon, Professor, Computer Science, Clemson University
• Kakan Dey, Assistant Professor, Civil Engineering, Clemson University
• Linh Bao Ngo, Director of Data Science, CITI/ACI-REF, Clemson University
• Ashwin Shrinath, Research Facilitator, CITI/ACI-REF, Clemson University
• Yuheng Du, Ph.D. Candidate, Computer Science, Clemson University
• Mizan Rahman, Ph.D. Candidate, Civil Engineering, Clemson University

Resources:

• Clemson University’s Palmetto Supercomputer
• Clemson University’s Holocron Cluster
• CloudLab

Funding Sources

The work described in this case study was supported in part by a grant from the National Science Foundation, Award #1341935, Advanced Cyberinfrastructure – Research and Educational Facilitation: Campus-Based Computational Research Support and the U.S. Department of Transportation Southeastern Transportation Center.