GRASP Lab faculty Annie Shera, Dinesh Jayaraman, Pratik Chaudhari, Eric Eaton, Nadia Figueroa, Lingjie Liu, and Antonio Loquercio, alongside students Elliot and Zach, present research overviews at Penn’s GRASP Lab.
Annie Shera’s group studies multi-robot systems affected by water and air flows, including underwater vehicles and quadrotors. Examples include flying quadrotors in close formation despite downwash and using robots to trace contaminants through buildings with changing HVAC flows. The group also studies water quality on the Schuylkill River with Earth and Environmental Sciences faculty and local organizations.
Dinesh Jayaraman argues that robot learning can draw on simulated physics and combine data from separate robotics components, rather than relying on a single massive dataset. His group uses coding agents to generate reward functions and training environments, design manipulation tools, choose useful scene representations, and coordinate robot software modules. Examples include training a quadruped to walk on a yoga ball and navigate obstacle courses.
Pratik Chaudhari’s group combines deep learning research with outdoor autonomy for forests, farms, and construction sites. Its NeuroFly drones use event cameras for fast perception in darkness, while the OctoSense model learns from data collected across robots, boats, and cars. Eric Eaton’s Lifelong Machine Learning Group develops adaptable systems, including robot teams for the DARPA Triage Challenge that locate casualties and report injuries to responders.
Nadia Figueroa’s lab learns stable robot behaviors from demonstrations and adapts them to safety constraints and human interaction. Applications include compliant manipulation, catching objects, stroke rehabilitation, and an exoskeleton guided by EMG signals and a musculoskeletal digital twin. Lingjie Liu’s lab reconstructs 3D geometry and physical properties from images and video, with projects spanning interactive scenes, humanoid control, VR teleoperation, and tactile manipulation.
Antonio Loquercio studies agile robots that can act quickly despite uncertainty, including a drone that improves its flight policy through repeated attempts. Elliot presents Nick Matni’s group’s work combining control theory and machine learning to improve reliability under model mismatch, perception errors, and distribution shift. Zach describes George Pappas’s research on robot foundation-model safety, including RoboPair experiments that elicited harmful behavior from a Unitree Go robot and other systems.
Briefs summarise talks and interviews; timestamps open the video at that moment. Claims in a talk are the speaker's own.
A GRASP Lab seminar host and panelists Nadia Figueroa, Dinesh Jayaraman, and Mark Yim discuss robotics research, deployment challenges, and future opportunities at the University of Pennsylvania.
Dr. Dinesh Jayaraman, a faculty member in computer and information science at the University of Pennsylvania and GRASP Lab, discusses visual robot learning and ways to make robots more capable in uncertain settings.
Dr. Nadia Figueroa, a new GRASP faculty member at the University of Pennsylvania, presents her research on collaborative, human-aware robotics and the challenges of making robots learn, adapt, and work safely with people.
