January 16, 2025 |

Revolutionary and Responsible AI Robotics at the TIER Lab

By Emily Blackburne

In the fast-growing field of artificial intelligence (AI) and robotics, Hunter College’s Trustworthy, Intelligent, and Explainable Robotics (TIER) Lab, is making significant strides in developing AI technology alongside responsible ethical considerations. Led by Hunter College assistant professor of Computer Science, Dr. Raj Korpan, the lab focuses on creating AI-powered robotic systems that not only excel in functionality but also serve society in ethical and equitable ways.

The TIER lab consists of Hunter college undergraduate students and a doctoral student, Daniel Weiner at the CUNYGraduate Center focused on robotics and machine learning. The lab’s mission addresses several critical ethical issues in AI such as human-robot interaction (HRI), bias in AI algorithms, representing diverse populations, and the formation of ethical guidelines for AI behavior in society.

“You want AI systems you can trust to act in an intelligent but explainable way, providing transparency and accountability� Korpan emphasizes.

Significant efforts of the lab focus on robot navigation in complex human environments, with undergraduate students Eric Guan, Georgina Woo, and Daniel Merino researching this. By developing deep machine learning AI algorithms that are also based on social cognition studies, the lab aims to enhance the way robots interact with people, ensuring these interactions are convenient and socially beneficial. Working alongside Hunter professor Dr. Susan Epstein and BMCC professor Dr. Mohammad Azhar, Korpan’s team tests these AI algorithms on detailed floor maps of the CUNYGraduate Center building. Doing so, they address challenges such as how to program task robots to best navigate spaces in a way that doesn’t make people feel uncomfortable, or violate personal space, while also ensuring the task gets done efficiently.

Along with Epstein and undergraduate students Daniel Tiorine and Sami Chen, Korpan researches explainable and transparent robotic systems that allows users to understand and trust why AI makes certain decisions. Undergraduate students Kelly Lin and Leah Kelley research this in relation to robot-human interaction and blame. This addresses who is accountable if a robot fails at a task or in providing guidance. For instance, if your train breaks down and an AI robot is there to guide you, but you’re late to work after taking its advice—who gets the blame? Using social cognition theories combined with analyzing body language and facial cues, students analyze human-human interactions, and program AI algorithms based on this, so that robot behavior is respectful and can intuitively adjust behavior to all users.

Korpan emphasizes the potential that CUNYespecially has in addressing ethical challenges that the private sector may overlook, exemplified by collaborations done across CUNYschools and the diverse populations that cross-discipline research is done with. “Academia provides an opportunity to create systems that aren’t profit-driven but community-driven,� he says. This ethos manifests in projects like participatory design with marginalized groups, where the lab collaborates directly with communities to ensure the systems they build would be beneficial. Undergraduate students in Korpan’s spring 2024 capstone course, Kevin Feng, Steven Hsui, Tiffanie Ku, and Peng Jie Wu, applied this ethos by noticing the need for more community and created the smartphone AI app “Linguabot� that helps users learn English using AI-powered conversation partners. Other undergraduate students in the course, Wei Da Chen, Daniel Maxime, Amy Ta, and Uriel Trejo developed the web app “WriteVibe� for neurodivergent users, which allows different animations, sound effects, and music alongside writing, to keep users engaged and focused on writing.

Along with serving community needs, the lab’s robots would also represent them. “We want robots to reflect the diversity of the communities they serve,� Korpan says. Through his leadership in the emerging field of responsible robotics, he is a core organizer of Queer in AI, and co-founder of Queer in Robotics—organizations dedicated to ensuring that AI and robotics systems serve marginalized LGBTQ+ AI users and scientists, who have been harmed by conventional AI robotic systems. Korpan proposes diversifying research participants, getting more participation and input from queer identifying people in HRI research, and redefining HRI trust through a queer perspective.

Another means of the lab’s representation in robotics is through exploring robot design and functionality. For instance, Korpan’s team explores how societal prejudices—like gender and racial biases—translate to interactions with humanoid or human-looking robots. This research highlights the importance of building robots that use inclusive language and are designed with diverse communities in mind, to prevent perpetuating stereotypes and bias.

Korpan emphasizes that ethical AI development relies on public policies prioritizing communities over profit-driven interests. Working alongside College of Staten Island professor Dr. Kathleen Cumiskey, who is also Director of the Public Interest Technology (PIT) initiative at ³Ô¹ÏÊÓÆµ, both advance civil technology and design curricula that combine public interests with technological advancements.

Korpan aims to expand opportunities for all students interested in pursuing careers in robotics and AI. He hopes to establish a partnership with Dr. Azhar and create a continuous pipeline between BMCC and Hunter College, allowing students to continue studies in AI and robotics from junior college through senior-level programs.

TIER Lab embraces the AI era, while emphasizing the need to address social issues. “People shouldn’t have to change themselves in order to accommodate these new systems. People shouldn’t change their behavior to accommodate a robot, the robot should adapt its behavior to conform to people’s needs and preferences — we’re ultimately building systems to support and assist people.â€� Korpan asserts. Truly revolutionary AI prioritizes human nature and needs in its design