About

Sylvia Herbert is a robotics and control researcher whose work sits at the intersection of safe autonomy, motion planning, and human-robot interaction. She is best known for developing principled frameworks that enable autonomous systems to operate safely in dynamic, uncertain environments — particularly in the presence of unpredictable human agents. Herbert's most influential contribution, "Confidence-aware motion prediction for real-time collision avoidance" (2019, 115 citations), introduced a novel approach to robot navigation that explicitly accounts for uncertainty in human motion predictions, a longstanding bottleneck in real-world deployment. This work laid groundwork for a broader research thread exploring probabilistically safe planning and confidence-based human prediction models, advancing the field's ability to handle the irreducible complexity of human behavior. Her more recent work bridges data-driven learning with formal safety tools, notably Control Barrier Functions and Hamilton-Jacobi reachability theory. Papers such as "Refining Control Barrier Functions through Hamilton-Jacobi Reachability" (2022) and "Sequential Neural Barriers for Scalable Dynamic Obstacle Avoidance" (2023) reflect her drive to make safety guarantees both theoretically rigorous and computationally scalable. Herbert's research addresses one of autonomy's most pressing open challenges: building robots that are not merely capable, but provably safe in the messy complexity of the real world.

Research Focus

Key Achievements

5
H-Index
10
Papers
197
Total Citations
20
Avg Citations/Paper
🏆 Most Cited Paper
Confidence-aware motion prediction for real-time collision avoidance <sup>1</sup>
115 citations · 2019
📈 Most Prolific Year: 2018 (3 Papers)
🤝 Key Collaborators: 22
🏛 Institutions: University of California, Berkeley, University of California San Diego, King Abdullah University of Science and Technology

Top Papers

  1. 1
  2. 2
  3. 3
  4. 4
  5. 5
  6. 6
  7. 7
  8. 8
  9. 9
  10. 10

Key Collaborators

Contact & Links

Available for collaboration
Content generated · 14 days ago