Geoff Fink

About

Geoff Fink is a leading researcher in legged robotics, specializing in locomotion control and state estimation for quadruped robots operating in challenging, unstructured environments. His work bridges the critical gap between rigid-terrain assumptions and the complex realities of soft, deformable ground. Fink’s most impactful contribution is the "STANCE" framework (68 citations), which addresses the failure of Whole-Body Control (WBC) on soft terrain by accounting for unmodeled contact dynamics, enabling more robust locomotion adaptation. He has also pioneered advanced proprioceptive sensor fusion techniques, developing low-level state estimators that integrate kinematics, IMU, LiDAR, and GPS data to achieve accurate attitude, odometry, and contact detection. His 2020 paper on proprioceptive sensor fusion (41 citations) and subsequent work on slip detection (15 citations) have become foundational for quadruped navigation. More recently, Fink introduced the MUSE estimator and invariant filtering frameworks (2025), which dramatically reduce position drift in real-time. His research is essential for enabling autonomous robots in search and rescue, inspection, and maintenance applications, where reliable perception and control over unpredictable terrain are paramount.

Research Focus

Key Achievements

Top Papers

1. 1
   STANCE: Locomotion Adaptation Over Soft Terrain

   68 citations · 2020
2. 2
   Proprioceptive Sensor Fusion for Quadruped Robot State Estimation

   41 citations · 2020
3. 3
   On Slip Detection for Quadruped Robots

   15 citations · 2022
4. 4
   Multi-Sensor Fusion for Quadruped Robot State Estimation Using Invariant Filtering and Smoothing

   5 citations · 2025
5. 5
   MUSE: A Real-Time Multi-Sensor State Estimator for Quadruped Robots

   4 citations · 2025
6. 6
   On State Estimation for Legged Locomotion Over Soft Terrain

   2 citations · 2021

Key Collaborators

Contact & Links