Research
Privacy-Sensitive Robot Design
My research examines how robots that share our private spaces can safely handle the information they collect and grant greater control of privacy to people. Here are highlights from the three primary threads of my work in the People and Robots Lab:
About This Research
As robots enter homes and other private settings, they may see, hear, and infer information people want to remain private. This work develops signal and inhibition mechanisms for data collection, explores LLM-enabled recognition of and response to private moments, and proposes approaches to limiting the disclosure and exposure of sensitive information to others.
Selected Papers
- See No Evil, Hear No Evil: Exploring Privacy-Preserving Designs for Social Robot Data Collection Conference manuscript under review
- Benchmarking LLM Privacy Recognition for Social Robot Decision Making Conference manuscript under review
- CONFIDANT: A Privacy Controller for Social Robots HRI 2022
- Examining the Effectiveness of Obfuscatory Planning Strategies through Human Observation HRI 2026 Companion · Honorable Mention, Best LBR 🏅
Other Research Highlights
Works on human-robot interaction beyond privacy
About This Research
Collaborative robots (cobots) can complement existing human workflows, but realizing this potential depends on how they are integrated and who is able to use them. This work helps stakeholders consider worker preferences, business goals, and cobot constraints as well as train novice operators to program cobots the way experts do, and considers what purposeful integration of robots in the real world requires.
Selected Papers
- Making Informed Decisions: Supporting Cobot Integration Considering Business and Worker Preferences HRI 2024
- Robots in the Real World: Promise, Peril, and Purposeful Integration XRDS: Crossroads, The ACM Magazine for Students, 2026
- CoFrame: A System for Training Novice Cobot Programmers Journal manuscript under review
About This Research
How robots behave, interact, and initially introduce themselves to the world around them is what helps truly bring them to life in the eyes of people. This work proposes the use of primal world beliefs to derive robot behavior and character from a consistent internal state, introduces Lively, a framework for real-time, lifelike robot motion, and explores how the unboxing process can imbue robots with character from their very first interaction.
Selected Papers
- Robot Primals: Exploring World Beliefs as a Source for Robot Behavior Design ACM Transactions on Human-Robot Interaction, 2026
- Lively: Enabling Multimodal, Lifelike, and Extensible Real-time Robot Motion HRI 2023 · Best Systems Paper 🏆
- Demonstrating the Potential of Interactive Product Packaging for Enriching Human-Robot Interaction HRI 2023 Companion