{"684608":{"#nid":"684608","#data":{"type":"event","title":"PhD Defense by Varun Agrawal","body":[{"value":"\u003Cp\u003E\u003Cstrong\u003ETitle\u003C\/strong\u003E: Proprioceptive State Estimation for Legged Robots with Probabilistic and Hybrid Kinodynamics\u003Cbr\u003E\u003Cbr\u003E\u003Cstrong\u003EVarun Agrawal\u003C\/strong\u003E\u003Cbr\u003EPh.D. Candidate in Robotics\u003Cbr\u003ESchool of Interactive Computing\u003Cbr\u003EGeorgia Institute of Technology\u003Cbr\u003E\u003Ca href=\u0022https:\/\/varunagrawal.github.io\/\u0022\u003Ehttps:\/\/varunagrawal.github.io\/\u003C\/a\u003E\u003Cbr\u003E\u003Cbr\u003E\u003Cstrong\u003EDate\u003C\/strong\u003E: Wednesday, September 17th, 2025\u003C\/p\u003E\u003Cp\u003E\u003Cstrong\u003ETime\u003C\/strong\u003E:\u003Cstrong\u003E\u0026nbsp;\u003C\/strong\u003E9:30 AM - 11:30 AM EST\u003Cbr\u003E\u003Cstrong\u003ELocation\u003C\/strong\u003E: Klaus Advanced Computing Building 2100 (\u003Ca href=\u0022https:\/\/maps.app.goo.gl\/1EYpk7UawxM4t6bP6\u0022\u003Emap\u003C\/a\u003E)\u003C\/p\u003E\u003Cp\u003E\u003Cstrong\u003EVirtual\u003C\/strong\u003E: LINK ON REQUEST\u003C\/p\u003E\u003Cp\u003E\u003Cstrong\u003ECommittee\u003C\/strong\u003E\u003C\/p\u003E\u003Cp\u003EDr. Frank Dellaert (Advisor),\u0026nbsp;School of Interactive Computing, Georgia Institute of Technology\u003C\/p\u003E\u003Cp\u003EDr. Sehoon Ha,\u0026nbsp;School of Interactive Computing, Georgia Institute of Technology\u003C\/p\u003E\u003Cp\u003EDr. Seth Hutchinson, Khoury College of Computer Sciences, Northeastern University\u003C\/p\u003E\u003Cp\u003EDr. Maurice Fallon, Oxford Robotics Institute, University of Oxford\u003C\/p\u003E\u003Cp\u003EDr. Robert Griffin, Institute for Human Machine Cognition \u0026amp; University of West Florida\u003C\/p\u003E\u003Cp\u003E\u003Cbr\u003E\u003Cstrong\u003ESummary\u003C\/strong\u003E\u003C\/p\u003E\u003Cp\u003ELegged robots hold immense promise for robot navigation within both human-centric and unstructured environments, due to their ability to perform dynamic motions, giving them high potential for widespread use and a myriad of applications. To be effective though, legged robot controllers require estimates of the robot\u0027s state at significantly high frequencies in order to plan and execute dynamic motions. Exteroceptive sensors, such as cameras and LiDAR, are unable to provide measurements at these desired rates, thus becoming a limiting factor. On the other hand, proprioceptive sensors, such as the Inertial Measurement Unit (IMU), are capable of providing measurements at immense frequencies and are not affected by changing environmental factors. However, low cost versions of these sensors are encumbered with significant and varied types of noise, yielding poor results when used directly.\u003C\/p\u003E\u003Cp\u003E\u0026nbsp;\u003C\/p\u003E\u003Cp\u003EIn this thesis, we propose algorithms and techniques for improving state estimation of legged robots using only proprioceptive sensing. We leverage and improve upon existing works on legged robot state estimation and probabilistic modeling to tackle various sources of noise and inaccuracies, improving the accuracy and robustness of the state estimates.\u003C\/p\u003E\u003Cp\u003E\u0026nbsp;\u003C\/p\u003E\u003Cp\u003EOur first contribution is a novel factor graph model for performing smoothing-based \u003Cem\u003EMaximum A Posteriori\u003C\/em\u003E\u0026nbsp;estimation, which takes advantage of the legged robot form factor. Building upon IMU preintegration theory and the use of environmental contact during leg stance phases for constraining sensor noise, we probabilistically model the kinematic chain of each leg within the robot, which compensates for nonlinear deformations and leads to improved accuracy. Furthermore, we show how leveraging M-estimators allows for automatic slip rejection, providing our estimator with greater robustness.\u003C\/p\u003E\u003Cp\u003E\u0026nbsp;\u003C\/p\u003E\u003Cp\u003EWe then introduce a new group-theoretic metric for measuring the performance of state estimators. Our metric combines the evaluation of the pose and the unobservable linear velocity of the robot into a singular form, providing a unified way to measure state estimator performance. This contrasts with existing metrics which measure the accuracy of the pose and the velocity estimates individually, thus leading to potential trade-offs and, consequently, subpar results. We further demonstrate the use of Chebyshev polynomial-based differentiation-via-interpolation to compute the true velocity from ground truth pose values, alleviating the need for expensive systems to measure the ground truth linear velocity.\u003C\/p\u003E\u003Cp\u003E\u0026nbsp;\u003C\/p\u003E\u003Cp\u003EOur third contribution is the development of a hybrid factor graph framework for modeling discrete and continuous estimation problems simultaneously, and a novel variable elimination algorithm for converting the hybrid factor graph into a hybrid Bayesian network. The proposed elimination algorithm results in exact posterior probabilities, overcoming the drawbacks of existing approximation based approaches. We further propose novel methods for constraining the computational and runtime complexities, making large scale deployment of our framework viable. This is demonstrated through the development of a hybrid legged robot state estimator which is capable of simultaneously estimating continuous robot states and discrete leg contact events using only kinematic information and measurements. We present results on both simulated robots and real-world hardware, showcasing the efficacy of our hybrid factor graph framework.\u003C\/p\u003E","summary":"","format":"limited_html"}],"field_subtitle":"","field_summary":[{"value":"\u003Cp\u003EProprioceptive State Estimation for Legged Robots with Probabilistic and Hybrid Kinodynamics\u003C\/p\u003E","format":"limited_html"}],"field_summary_sentence":[{"value":"Proprioceptive State Estimation for Legged Robots with Probabilistic and Hybrid Kinodynamics"}],"uid":"27707","created_gmt":"2025-09-08 12:50:37","changed_gmt":"2025-09-08 12:51:46","author":"Tatianna Richardson","boilerplate_text":"","field_publication":"","field_article_url":"","field_event_time":{"event_time_start":"2025-09-17T09:30:03-04:00","event_time_end":"2025-09-17T11:30:00-04:00","event_time_end_last":"2025-09-17T11:30:00-04:00","gmt_time_start":"2025-09-17 13:30:03","gmt_time_end":"2025-09-17 15:30:00","gmt_time_end_last":"2025-09-17 15:30:00","rrule":null,"timezone":"America\/New_York"},"location":"Klaus Advanced Computing Building ","extras":[],"groups":[{"id":"221981","name":"Graduate Studies"}],"categories":[],"keywords":[{"id":"100811","name":"Phd Defense"}],"core_research_areas":[],"news_room_topics":[],"event_categories":[{"id":"1788","name":"Other\/Miscellaneous"}],"invited_audience":[{"id":"78771","name":"Public"}],"affiliations":[],"classification":[],"areas_of_expertise":[],"news_and_recent_appearances":[],"phone":[],"contact":[],"email":[],"slides":[],"orientation":[],"userdata":""}}}