Sarah Keren

About

Sarah Keren is a leading researcher at the intersection of artificial intelligence, robotics, and human-agent interaction, whose work fundamentally rethinks how we design environments to make intelligent agents—both robots and humans—more effective and interpretable. Her most influential contribution is the pioneering framework of **Goal Recognition Design (GRD)** , which she introduced to proactively analyze and redesign environments so that an observer can quickly and reliably infer an agent’s objective. This work, culminating in a comprehensive 2020 survey, has garnered over 30 citations and established a new subfield in AI. Keren also developed the **Equi-Reward Utility Maximizing Design (ER-UMD)** model, which tackles the offline redesign of stochastic environments to maximize agent performance—a concept with direct applications in human-robot collaboration and smart infrastructure. Her research on designing environments for interpretable robot behavior has been recognized as essential for building trust in autonomous systems. With a growing citation record exceeding 60 total citations, Keren’s work is shaping how we build spaces that are not just functional, but inherently legible and cooperative for the agents that inhabit them.

Research Focus

Key Achievements

Top Papers

1. 1
   Equi-Reward Utility Maximizing Design in Stochastic Environments

   22 citations · 2017
2. 2
   Goal Recognition Design in Deterministic Environments

   19 citations · 2019
3. 3
   Goal Recognition Design - Survey

   10 citations · 2020
4. 4
   Designing Environments Conducive to Interpretable Robot Behavior

   5 citations · 2020
5. 5
   Redesigning Stochastic Environments for Maximized Utility

   4 citations · 2017
6. 6
   Value of Assistance for Mobile Agents

   2 citations · 2023

Key Collaborators

Contact & Links