Research

Publications and Projects

Robotics · Artificial Intelligence · Engineering Education

IEEE SoutheastCon 2026

Large Language Models for Nodal Analysis in Circuits Education: An Evaluation

Yue Cao · IEEE SoutheastCon · 2026 Huntsville AL

Evaluations of the state-of-the-art LLMs in solving undergraduate-level Nodal Analysis tasks.

ACM/IEEE SEC 2025

Performance Evaluation of Whisper-Series Speech Transcription Models on Raspberry Pi

Yue Cao · ACM/IEEE Symposium on Edge Computing - Edge Intelligence Workshop · 2025 Arlington VA

Benchmarking Whisper & Faster Whisper models on Raspberry Pi platforms, including latency, memory usage, and thermal behavior.

Ph.D. Dissertation 2024

Towards Manipulator Task-Oriented Programming: Automating Behavior-Tree Configuration

Yue Cao · Purdue University, ECE, West Lafayette · 2024

Task-Oriented Programming: Program manipulators in terms of high-level tasks, instead of in terms of explicit motions.

Also as Poster at the Northeast Robotics Colloquium, Ithaca NY, 2025.

AAAI Symposium 2023

Ground Manipulator Primitive Tasks to Executable Actions using Large Language Models

Yue Cao and C. S. George Lee · AAAI Fall Symposium on Unifying Representations for Robot Application Development · 2023 Arlington VA

Enabling LLMs to generate position/force set-points for manipulator hybrid control.

Also as Late-Breaking Results at IROS, Detroit MI, 2023.

AAAI Symposium 2023

Robot Behavior-Tree-Based Task Generation with Large Language Models

Yue Cao and C. S. George Lee · AAAI Spring Symposium on Challenges Requiring the Combination of Machine Learning and Knowledge Engineerin · 2023 San Francisco CA

Enabling a hierarchical-structured robot task generation. Three key components: Phase-Step prompt, behavior-tree construction, and automatic source-task selection

IROS 2022

Behavior-Tree Embeddings for Robot Task-Level Knowledge

Yue Cao and C. S. George Lee · IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS) · 2022 Kyoto Japan

Converting symbolic task knowledge to numerical form.
A new approach to reuse task knowledge.
Enabling machine learning on symbolic tasks.

×