By Neil Nachbar
In the fall semester of 2018, a new Artificial Intelligence (AI) Lab will be accessible to all University of Rhode Island students on the first floor of the Robert L. Carothers Library. Funded by a $143,065 grant from The Champlin Foundation, the AI Lab is believed to be the first in the nation to be located in a library.
The primary principal investigator of the lab is Dr. Kunal Mankodiya, assistant professor in URI’s Department of Electrical, Computer and Biomedical Engineering. Mankodiya established a founding team of the AI lab, involving the URI College of Arts and Sciences, and URI Libraries. Other founding members are Cheryl Foster, Department of Philosophy; Joan Peckham, Department of Computer Science and Statistics/Big Data Initiative and Data Science Programs; and Harrison Dekker, Angel Ferria, and Karim Boughida, all of University Libraries.
“The popularity of AI and associated fields is not just limited to industries,” said Mankodiya. “Our team looked at a recent survey asking URI students about topics they wished to see in their curriculum. AI was among the top requests. This prompted us to initiate thoughts on the AI lab.”
The AI Lab will serve as a collaborative environment for students who are working on their capstone or class projects. Students will be required to register themselves and go through various trainings, from basics to advanced AI tutorials, which will prepare them to apply AI in their projects.
“We have worked out a plan to invite other faculty to leverage this facility in their own curriculum,” said Mankodiya. “Many times, the students have less resources to execute their projects. In this case, the AI Lab becomes an important venue to engage and convert their ideas into things. In other words, the AI Lab will enable students to make things that can think.”
Since joining the College of Engineering faculty, Mankodiya and his students have worked on many projects that involve technology that relates, at least indirectly, to AI.
“Although I have not dealt directly with AI, I have gained a broad experience in embedded intelligence where the modern technologies such as Internet of Things (IoT), fog/edge computing, and deep learning are moving further,” stated the professor. “I have experience developing medical IoT technologies involving wearable devices, embedded intelligence, and big data. In these technologies, my research team continuously design artificial intelligent algorithms that can predict the progression of disorders such as Parkinson’s disease. In addition, I was fortunate to have found enthusiastic members from URI who shared and supported the vision on the AI Lab.”
Considering the broad appeal of the AI Lab, the Carothers Library made the most sense for a location.
“Our AI Lab team found the URI library as an easy-to-access facility at a centralized location, which not only is neutral among disciplines, but also serves thousands of information-seeking students per day,” said Mankodiya.
Once the lab is operational, the lab team plans on coordinating some events to draw attention and interest among faculty and students.
“After the establishment of the lab, we will create workshops and hackathons to promote the AI Lab so that other faculty feel invited to use the space,” said Mankodiya.
The AI Lab will occupy a 600-square-foot room, which will have three zones.
Zone 1: AI Workstations for Individualized Learning
The lab will contain eight identical Lambda workstations where a student or a student team can learn about AI and relevant subject areas. Each workstation is based on a supercomputer where students, beginning with introductory exercises, can easily transition into hands-on projects in the next level.
Zone 2: Hands-on Project Bench
After students receive basic training on AI and data science, they move to advanced tools to design hands-on projects in which they can apply AI algorithms to various applications:
- Deep Learning Robots: An environment will be created for students to program mobile robots instrumented with cameras, radars, sensors, and actuators. A line maze right outside the lab will be deployed for the mobile robots. Students will build AI algorithms to navigate both known and unknown environments.
- Internet-of-things for Smart Cities: A physical model of Rhode Island will be developed. The model will contain distributed sensors and actuators so students can design intelligent algorithmic projects to control city environments such as lighting, traffic, transit and parking.
- Big Data Analytics: A learning environment will be designed with other faculty so students can access and analyze various kinds of big data available in public domain.
Zone 3: AI Hub for Collaborative Thinking
In the center of the AI Lab, an AI hub will be located where students with experts can pursue collaborative thinking including design thinking and brainstorming about AI beyond design.