BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Robotics Institute Carnegie Mellon University - ECPv6.15.12.1//NONSGML v1.0//EN
CALSCALE:GREGORIAN
METHOD:PUBLISH
X-WR-CALNAME:Robotics Institute Carnegie Mellon University
X-ORIGINAL-URL:https://www.ri.cmu.edu
X-WR-CALDESC:Events for Robotics Institute Carnegie Mellon University
REFRESH-INTERVAL;VALUE=DURATION:PT1H
X-Robots-Tag:noindex
X-PUBLISHED-TTL:PT1H
BEGIN:VTIMEZONE
TZID:America/New_York
BEGIN:DAYLIGHT
TZOFFSETFROM:-0500
TZOFFSETTO:-0400
TZNAME:EDT
DTSTART:20250309T070000
END:DAYLIGHT
BEGIN:STANDARD
TZOFFSETFROM:-0400
TZOFFSETTO:-0500
TZNAME:EST
DTSTART:20251102T060000
END:STANDARD
BEGIN:DAYLIGHT
TZOFFSETFROM:-0500
TZOFFSETTO:-0400
TZNAME:EDT
DTSTART:20260308T070000
END:DAYLIGHT
BEGIN:STANDARD
TZOFFSETFROM:-0400
TZOFFSETTO:-0500
TZNAME:EST
DTSTART:20261101T060000
END:STANDARD
BEGIN:DAYLIGHT
TZOFFSETFROM:-0500
TZOFFSETTO:-0400
TZNAME:EDT
DTSTART:20270314T070000
END:DAYLIGHT
BEGIN:STANDARD
TZOFFSETFROM:-0400
TZOFFSETTO:-0500
TZNAME:EST
DTSTART:20271107T060000
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
DTSTART;TZID=America/New_York:20260526T133000
DTEND;TZID=America/New_York:20260526T143000
DTSTAMP:20260707T165816
CREATED:20260520T151822Z
LAST-MODIFIED:20260520T151822Z
UID:151376-1779802200-1779805800@www.ri.cmu.edu
SUMMARY:Explore and Exploit: Learning Policies for Efficient and Coordinated Active Search
DESCRIPTION:Abstract:  \nRobotic search is becoming a central capability in domains where the world is large\, uncertain\, and costly to inspect directly: search and rescue\, environmental monitoring\, surveillance\, and infrastructure inspection. In these settings\, the hard problem is not perception alone but the online sensing decision: where to look next as evidence arrives\, while every motion command spends travel distance and mission time. Active search requires both exploration and exploitation: reducing uncertainty across the searchable region\, and using the emerging posterior to confirm likely objects of interest. This thesis studies that decision problem under a shared Bayesian active-search formulation\, and develops learned policies for real-time field execution and coordination across heterogeneous teams. \nThe first contribution is a path-aware single-robot policy. We introduce a path-integral expert that scores complete shortest-path routes by their accumulated expected information gain\, then amortize that expensive expert into a Graph Attention Network policy by behavior cloning. The resulting policy achieves up to a 254x speedup over the path-integral expert and an 11x speedup over a one-step greedy expert\, generalizes across map geometries and object densities without retraining\, and yields up to 2.86x faster mission completion against a strong NATS baseline in field tests on an autonomous ground vehicle in a 75\,000 square meter forested environment. \nThe second contribution extends the same shared-belief interface from single-robot search to coordinated teams. We propose QUEST\, a multi-agent Q-learning framework that trains over the same graph backbone but optimizes a per-decision Bellman target over downstream belief\, robot positions\, team coverage\, and remaining budget. The formulation covers single robots\, homogeneous UGV teams\, and heterogeneous UAV-UGV teams. QUEST reaches 0.913 F-score with four UGVs versus 0.853 for the strongest learned baseline\, maintains its lead under communication outage and two simultaneous mid-mission robot failures without retraining\, and on out-of-distribution UAV-UGV compositions improves early F-score\, 0.759 versus 0.728\, while using 10% less path length and 12% less duplicate coverage. A map-structure analysis using algebraic connectivity and modularity ties these gains to the topology of the search graph\, predicting when non-myopic value learning is and is not worth its training cost. \nTogether\, these policies connect field deployment with coordinated team search that generalizes across composition\, communication\, and partial failure under one shared-belief active-search formulation. \n\nCommittee:\nDr. Jeff Schneider (chair)\nDr. Wennie Tabib\nDr. Tejus Gupta
URL:https://www.ri.cmu.edu/event/explore-and-exploit-learning-policies-for-efficient-and-coordinated-active-search/
LOCATION:Newell-Simon Hall 4305
CATEGORIES:MSR Thesis Presentation,Student Talks
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=America/New_York:20260526T103000
DTEND;TZID=America/New_York:20260526T113000
DTSTAMP:20260707T165816
CREATED:20260520T151702Z
LAST-MODIFIED:20260520T151702Z
UID:151374-1779791400-1779795000@www.ri.cmu.edu
SUMMARY:Toward Real-World Autonomous Off-Road Driving with Reinforcement Learning
DESCRIPTION:Abstract:\nOff-road autonomous driving presents significant challenges such as navigating unmapped variable environments\, traversing difficult terrain geometries such as steep slopes and ditches\, and managing complex terrain dynamics. Addressing these challenges requires effective low-level adaptable control and long-horizon planning. Most existing methods utilize Model Predictive Control (MPC) methods such as Model Predictive Path Integral (MPPI)\, which have long-horizon planning capabilities\, but require expensive dense sampling and precise dynamics modeling that are impractical to deploy for real-time control. On the other hand\, Reinforcement Learning (RL) learns complex dynamics and reactive low-level control policies directly from interaction\, but typically fails to plan and navigate dense environments due to poor exploration.\n\nTo utilize the different strengths of both MPC methods and RL\, this thesis proposes a hierarchical pipeline with a low-frequency high-level MPPI planner for long-horizon planning and a high-frequency low-level adaptable RL controller. In order to solve RL’s poor exploration problem\, this thesis introduces Teacher Action Distillation Policy Optimization (TADPO)\, a novel teacher-student policy gradient formulation that extends Proximal Policy Optimization (PPO) to guide policy learning\, leveraging off-policy teacher trajectories for teacher guidance and on-policy trajectories for student exploration.\n\nWith TADPO\, we develop a vision-based\, end-to-end RL policy and off-road autonomy system for high-speed driving\, capable of navigating extreme slopes and obstacle-dense terrain. Firstly\, we demonstrate our performance in simulation and\, more importantly\, zero-shot sim-to-real transfer on a full-scale off-road vehicle. Then\, we show further performance improvements of the deployed policy in simulation. Finally\, we illustrate TADPO’s generalizability with a policy that navigates through trees in simulation. To our knowledge\, this work represents the first deployment of RL-based policies on a full-scale off-road robotics platform. \nCommittee:\nJeff Schneider (advisor)\nJohn Dolan\nSamuel Triest
URL:https://www.ri.cmu.edu/event/toward-real-world-autonomous-off-road-driving-with-reinforcement-learning/
LOCATION:Newell-Simon Hall 4305
CATEGORIES:MSR Thesis Presentation,Student Talks
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=America/New_York:20260522T153000
DTEND;TZID=America/New_York:20260522T170000
DTSTAMP:20260707T165816
CREATED:20260514T153637Z
LAST-MODIFIED:20260514T153734Z
UID:151265-1779463800-1779469200@www.ri.cmu.edu
SUMMARY:RI PhD Thesis Defense - Kenneth Shaw
DESCRIPTION:Date: Friday May 22\nTime: 3:30PM – 4:30PM (EST)\nLocation: NSH (Newell Simon Hall) 3305 \nZOOM Link \nTitle: Building Robot Hands and Teaching Dexterity \nAbstract:  \nOur human hands are masterpieces of power and precision\, capable of typing\, hammering\, or delicately using chopsticks. Yet most robots today still rely on simple two-finger grippers in controlled settings because dexterous hands are costly and difficult to deploy. To close this gap\, I will introduce my LEAP Hands\, high-performance\, low-cost\, and easy-to-assemble robotic hands that have become the most widely used platform for dexterous manipulation research. LEAP Hand V1 employs motor-in-joint actuation for simplicity\, while V2 introduces a novel hybrid rigid–soft structure that delivers exceptional strength and durability.  I will then show how large-scale human video/motion data and simulation techniques can teach human-like manipulation skills across diverse environments.  By tightly integrating mechanical design and machine learning\, my open-source robot hands achieve unprecedented levels of dexterity for a variety of everyday tasks.  To conclude\, I will discuss emerging directions in the field of dexterous manipulation research and share my future vision that I will pursue as a faculty member at the University of Illinois. \n​Prof. Deepak Pathak (advisor) \nProf. Nancy Pollard \nProf. Abhinav Gupta \nProf. Jitendra Malik\, UC Berkeley \nDr. Ankur Handa\, NVIDIA \n  \nThesis Draft
URL:https://www.ri.cmu.edu/event/ri-phd-thesis-defense-kenneth-shaw/
LOCATION:Newell-Simon Hall 3305
CATEGORIES:PhD Thesis Defense,Student Talks
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=America/New_York:20260521T153000
DTEND;TZID=America/New_York:20260521T170000
DTSTAMP:20260707T165816
CREATED:20260511T201231Z
LAST-MODIFIED:20260511T201231Z
UID:151239-1779377400-1779382800@www.ri.cmu.edu
SUMMARY:RI PhD Thesis Proposal - Anurag Ghosh
DESCRIPTION:Date: May 21st\, 2026\nTime: 3:30 – 5:00 pm\nRoom: NSH Room 4305\nZoom:  https://cmu.zoom.us/j/98318417145 \nType: RI PhD Thesis Proposal\nWho: Anurag Ghosh\n\n\nTitle: Scaling Long-Tailed Driving Perception and Planning with In-the-Wild Videos\n\n\nAbstract: Closed-loop driving\, where methods produce actions a simulator reacts to\, remains largely tied to driving logs from instrumented fleets. Thus\, reliably driving in rare-but-critical scenarios is still elusive. Meanwhile\, foundation models are increasingly common in autonomous driving\, vision-language and video world models are tackling open-loop tasks like scene description and video generation. Although the internet has revolutionized language and image generation\, planning in autonomous driving has not seen its ImageNet moment yet. Therefore\, an opportunity exists to leverage internet-scale data and tackle long-tailed autonomous driving.\n\nWe focus on work zones as a representative long-tail scenario as they are a major source of disengagements for commercial systems today. Work zones uniquely combine rare objects (e.g.\, construction vehicles\, arrow boards)\, unusual layouts (e.g.\, temporary closures\, crossing yellow lines)\, and unpredictable behaviors (e.g.\, flaggers\, sudden merges). These safety-critical scenarios are considered difficult to simulate at scale. \nFirst\, we mine long-tailed driving videos from a massive corpus. We find that foundation models fail at work zone perception and fine-tuning on our data combined with simple priors makes them effective. Second\, we introduce a resource-efficient\, geometry-based prior that improves scene perception and long-tail object detection. Third\, we focus on long-tailed closed-loop planning and develop an anytime language-action planner capable of real-time trajectory generation and contextual textual reasoning. Furthermore\, by developing a novel rules-based planner that effectively handles current benchmark scenarios\, we show that existing closed-loop driving benchmarks are insufficient for evaluating long-tailed behaviors. \nFinally\, this thesis proposes a framework that\, by carefully composing geometry-aware methods\, street-view imagery\, and foundation models\, lifts monocular videos into metric\, geo-referenced 4D driving logs compatible with existing simulators. Using this framework\, we create a new long-tail planning benchmark and propose to uncover insights and study the geographic scaling behavior of state-of-the-art planning methods. \nUltimately\, to advance autonomous driving beyond fleets\, we argue scaling of training and evaluation is achievable by harnessing internet-scale data while grounding foundation models with geometric and physical priors.\n\nCommittee:\nSrinivasa Narasimhan\, Chair\nDeva Ramanan\nMaxim Likhachev\nChristoph Mertz\nManmohan Chandraker\, UC San Diego\n\nThesis Proposal Draft
URL:https://www.ri.cmu.edu/event/ri-phd-thesis-proposal-anurag-ghosh/
LOCATION:Newell-Simon Hall 4305
CATEGORIES:PhD Thesis Proposal,Student Talks
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=America/New_York:20260521T123000
DTEND;TZID=America/New_York:20260521T133000
DTSTAMP:20260707T165816
CREATED:20260515T135220Z
LAST-MODIFIED:20260515T135220Z
UID:151275-1779366600-1779370200@www.ri.cmu.edu
SUMMARY:Generalizable Neural Dynamics Modeling for Complex Deformable Object Manipulation
DESCRIPTION:Abstract:\nHair care is an essential daily activity\, yet it remains inaccessible to individuals with limited mobility and challenging for autonomous robot systems due to the fine-grained physical structure and complex dynamics of hair.\nThis thesis presents DYMO-Hair\, a model-based robot hair care system. We introduce a novel dynamics learning paradigm that is suited for volumetric quantities such as hair\, relying on an action-conditioned latent state editing mechanism\, coupled with a compact 3D latent space of diverse hairstyles to improve generalizability. This latent space is pre-trained at scale using a novel hair physics simulator\, enabling generalization across previously unseen hairstyles. Using the dynamics model with a Model Predictive Path Integral (MPPI) planner\, DYMO-Hair is able to perform multimodal goal-conditioned hair styling\, supporting both visual references and natural language instructions as goal configurations.\nExperiments in simulation demonstrate that DYMO-Hair’s dynamics model outperforms baselines on capturing local deformation for diverse\, unseen hairstyles. DYMO-Hair further outperforms baselines in closed-loop visual goal-conditioned hair styling tasks on unseen hairstyles\, with an average of 22% lower final geometric error and 42% higher success rate than the state-of-the-art system. Real-world experiments exhibit zero-shot transferability of our system to wigs\, achieving consistent success on challenging unseen hairstyles where the state-of-the-art system fails. We further conduct a user study to evaluate the system’s language goal-conditioned styling capability\, suggesting improved goal achievement and progress consistency over zero-shot VLM-based methods.\nTogether\, these results introduce a foundation for model-based robot hair care\, advancing toward more generalizable\, flexible\, and accessible robot hair styling in unconstrained physical environments.\n\nCommittee:\nProf. Jean Oh (Chair)\nProf. Jeffrey Ichnowski\nProf. David Held\nUksang Yoo
URL:https://www.ri.cmu.edu/event/generalizable-neural-dynamics-modeling-for-complex-deformable-object-manipulation/
CATEGORIES:MSR Thesis Presentation,Student Talks
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=America/New_York:20260521T103000
DTEND;TZID=America/New_York:20260521T113000
DTSTAMP:20260707T165816
CREATED:20260518T135147Z
LAST-MODIFIED:20260518T135147Z
UID:151279-1779359400-1779363000@www.ri.cmu.edu
SUMMARY:Scalable Imitation Learning for Robust Manipulation and Physical Human-Robot Interaction
DESCRIPTION:Abstract:\nRobots in everyday human environments are expected to perform robust manipulation across cluttered\, constrained\, and physically interactive settings. This thesis studies how scalable simulation-based data generation can train vision-based imitation learning policies for robust zero-shot transfer to the real world. \nIn the first part of this thesis\, we study robotic manipulation in cluttered shelf environments\, where limited free space\, dense object arrangements\, and complex visual backgrounds make policy learning challenging. We present a scalable learning pipeline that leverages Dynamic Movement Primitives (DMPs) to expand a small set of teleoperated demonstrations into a large and diverse synthetic dataset. We train an imitation learning policy across six scenarios and three manipulation strategies\, demonstrating robust generalization across diverse object shapes and zero-shot transfer to a physical manipulator.\nIn the second part of this thesis\, we address autonomous physical human-robot interaction (pHRI)\, where large-scale real-world training data is difficult to collect. We introduce a zero-shot “text2sim2real” generative simulation framework that synthesizes diverse pHRI scenarios from high-level natural-language prompts\, including soft-body human models\, scene layouts\, and robot motion trajectories. Using the generated demonstrations\, we train vision-based imitation learning policies on segmented point clouds and show zero-shot sim-to-real transfer on scratching and bathing tasks\, with success rates exceeding 80%.\nTogether\, these two works demonstrate that scalable simulation data generation can serve as a practical foundation for robust robot policies in both constrained manipulation and physical human-robot interaction.\n\nCommittee:\nProf. Zackory Erickson\, chair\nProf. David Held\nYufei Wang
URL:https://www.ri.cmu.edu/event/scalable-imitation-learning-for-robust-manipulation-and-physical-human-robot-interaction/
CATEGORIES:MSR Thesis Presentation,Student Talks
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=America/New_York:20260520T150000
DTEND;TZID=America/New_York:20260520T163000
DTSTAMP:20260707T165816
CREATED:20260512T145730Z
LAST-MODIFIED:20260512T145730Z
UID:151245-1779289200-1779294600@www.ri.cmu.edu
SUMMARY:RI PhD Thesis Proposal - Renos Zabounidis
DESCRIPTION:Date: May 20th\, 2026\nTime: 3:00 – 4:30 pm\nLocation: NSH 4305\nZoom Link \nType: RI PhD Thesis Proposal\nWho: Renos Zabounidis \nTitle: Enforcing Neuro-Symbolic Structure in Deep Reinforcement Learning \nAbstract: Monolithic deep reinforcement learning trains a single network to learn vision\, physics\, planning\, and control from reward alone. The result is poor sample efficiency\, brittle generalization\, and uninterpretable decisions. This thesis shows how to build domain knowledge into policy architecture and enforce these architectural priors during training. \nWe develop this claim at three levels of abstractions. Concept-level abstractions route predictions through human-interpretable predicates such as `door_present’ and `key_in_inventory’\, enabling runtime inspection and intervention. Action-level constraints enforce state-dependent action validity\, preventing unmasked training from suppressing rarely valid behaviors through shared representations. Compositional abstractions represent long-horizon tasks as reusable symbolic skills that a planner can sequence while RL grounds each skill in low-level control. \nBuilding on these foundations\, this proposal focuses on two future directions. Concept-conditioned latent action models impose semantic structure on variational motor representations\, allowing high-level controllers to sample behaviors by name. A planning-guided option critic learns dynamic skill scheduling under precondition constraints\, replacing static plan traversal with on-policy option selection. \nTogether\, these contributions show that domain knowledge in the policy architecture reduces sample complexity\, enables cross-task skill composition\, and makes internal decisions available for inspection and override. \nThesis Committee:\nKatia Sycara\, CMU (Chair)\nSebastian Scherer\, CMU\nYonatan Bisk\, CMU\nKevin Ellis\, Cornell University \n\nThesis URL
URL:https://www.ri.cmu.edu/event/ri-phd-thesis-proposal-renos-zabounidis/
LOCATION:Newell-Simon Hall 4305
CATEGORIES:PhD Thesis Proposal,Student Talks
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=America/New_York:20260519T130000
DTEND;TZID=America/New_York:20260519T150000
DTSTAMP:20260707T165816
CREATED:20260514T173912Z
LAST-MODIFIED:20260514T173912Z
UID:151271-1779195600-1779202800@www.ri.cmu.edu
SUMMARY:RI PhD Thesis Defense - Junyu Nan
DESCRIPTION:Date: Tuesday May 19\, 2026\nTime:  1:00 – 3:00PM (EST)\nLocation: NSH (Newell Simon Hall) 3305 \nZoom Link \nType: PhD Thesis Defense \nTitle: Learning Geometric\, Physical\, and Semantic Priors for Embodied Planning and Control \nAbstract: Embodied intelligence requires perceiving\, predicting\, and acting in environments with an understanding of the geometric\, semantic\, and physical structure of the world. Recently\, the dominant trend in robotics has been to acquire such world understanding implicitly in a data-driven manner using end-to-end models. While these approaches have achieved impressive milestones\, they often rely on substantial amounts of in-domain data and may remain brittle when success depends on long-horizon reasoning\, precise physical interaction\, or generalization from limited task-specific data. This thesis studies a perspective complementary to end-to-end approaches: when aspects of geometric\, physical\, and semantic structure are known and reusable\, explicitly learning priors over them can improve the data efficiency\, fidelity\, and robustness of embodied learning systems. \nThis thesis instantiates this perspective across four embodied settings: scene-level forecasting in autonomous driving\, learning deformable object dynamics from robot interaction videos\, relational reasoning and cross-instance manipulation transfer\, and zero-shot long-horizon manipulation. For scene-level prediction\, we learn a predictive geometric prior over the future evolution of the full 3D scene representation. By modeling future motion at the scene level\, the predictive geometric representation preserves coherence across agents and the environment\, improving downstream prediction and planning under multi-agent uncertainty. Moving from passive scene forecasting to manipulation\, physical priors become important to predict how state evolves in response to robot actions. To learn contact-rich dynamics and topological change directly from RGB-D robot interaction videos\, we represent deformable objects as adaptive sets of 3D Gaussians inside a particle-filtering framework with physics-inspired interaction modeling and resampling mechanisms. Extending to multi-object manipulation problems\, we develop semantic priors based on correspondence as a reusable representation for relational reasoning and cross-instance manipulation. These correspondence-based priors allow a robot to identify functionally meaningful object structure\, reason about object alignment under geometric ambiguity\, and transfer manipulation knowledge from demonstrated objects to novel instances. Finally\, we integrate learned geometric\, semantic\, and physical priors into a zero-shot long-horizon manipulation system that connects high-level video and language planning with executable robot motion through geometric grounding of generated videos. Taken together\, these works show that explicitly learned geometric\, physical\, and semantic priors can improve the data efficiency\, fidelity\, and robustness of embodied prediction\, planning\, and manipulation systems. \nCommittee: \nKris Kitani (Chair)\, Carnegie Mellon University \nDavid Held\, Carnegie Mellon University \nShubham Tulsiani\, Carnegie Mellon University \nBrian Okorn\, Robotics and AI Institute \nThesis Draft
URL:https://www.ri.cmu.edu/event/ri-phd-thesis-defense-junyu-nan/
LOCATION:Newell-Simon Hall 3305
CATEGORIES:PhD Thesis Defense,Student Talks
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=America/New_York:20260518T130000
DTEND;TZID=America/New_York:20260518T150000
DTSTAMP:20260707T165816
CREATED:20260319T204042Z
LAST-MODIFIED:20260319T204042Z
UID:150640-1779109200-1779116400@www.ri.cmu.edu
SUMMARY:RI PhD Speaking Qualifier / MSR Talk - Yuemin Mao
DESCRIPTION:TBD
URL:https://www.ri.cmu.edu/event/ri-phd-speaking-qualifier-msr-talk-yuemin-mao/
LOCATION:Newell-Simon Hall 4305
CATEGORIES:PhD Speaking Qualifier,Student Talks
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=America/New_York:20260518T130000
DTEND;TZID=America/New_York:20260518T143000
DTSTAMP:20260707T165816
CREATED:20260514T143554Z
LAST-MODIFIED:20260514T143554Z
UID:151263-1779109200-1779114600@www.ri.cmu.edu
SUMMARY:Contact-Rich Manipulation with Vibro-Tactile Sensing
DESCRIPTION:Abstract: \nContact-rich manipulation requires reasoning about complex physical interactions. While vision and force sensing provide useful information\, many interaction dynamics remain difficult to observe directly\, especially under occlusion\, rapid contact transitions\, or distributed contact. This thesis explores how vibro-tactile sensing\, also known as structure-born acoustic sensing\, can serve as a practical and information-rich modality for understanding physical interaction in robotic manipulation. \nWe present three works that progressively study vibro-tactile sensing for contact-rich manipulation. First\, we present acoustic-guided constraint learning for fast non-prehensile transport\, where vibro-tactile sensing is used as a binary slip detector to learn motion-dependent friction effects\, and incorporate them into optimization-based motion planning. Second\, we present a multi-channel acoustic sensing system embedded in a parallel-jaw gripper\, which predicts continuous in-hand slip direction and magnitude in real time for closed-loop manipulation. Finally\, we introduce a visuo-acoustic framework that combines wearable active acoustic sensing with vision to estimate hand pose and contact during human-object interaction\, improving robustness under occlusion and visual ambiguity. \nTogether\, these works demonstrate that vibro-tactile sensing provides a scalable and effective modality for modeling\, representing\, and perceiving contact-rich interaction in robotic manipulation. \n  \nCommittee: \nProf. Jeffrey Ichnowski (Chair) \nProf. Christopher G. Atkeson \nProf. Maxim Likhachev \nYufei Wang
URL:https://www.ri.cmu.edu/event/contact-rich-manipulation-with-vibro-tactile-sensing/
LOCATION:Newell-Simon Hall 4305
CATEGORIES:MSR Thesis Presentation,Student Talks
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=America/New_York:20260518T110000
DTEND;TZID=America/New_York:20260518T123000
DTSTAMP:20260707T165816
CREATED:20260508T143910Z
LAST-MODIFIED:20260508T143910Z
UID:151231-1779102000-1779107400@www.ri.cmu.edu
SUMMARY:Designing A Learning-Enabled Non-Anthropomorphic Robotic Hand Framework for Dexterous Manipulation
DESCRIPTION:Abstract: As robots transition into unstructured human environments\, achieving dexterous manipulation across diverse tasks becomes increasingly crucial yet remains challenging. This objective underscores the value of high-degree-of-freedom (DoF) robotic hands capable of versatile grasping and in-hand manipulation. Complementing the hardware\, success in these settings also depends on intelligent and autonomous hand control strategies that provide the adaptability necessary for skill execution. \n\nIn this talk\, I first introduce DeltaHands\, a modular\, non-anthropomorphic hand framework. Its design provides a reconfigurable space that enables rapid adaptation of the hand’s DoFs\, mechanical properties\, material composition\, and sensor integration. While this non-anthropomorphic morphology offers mechanical simplicity\, its divergence from the human hand complicates intuitive human-in-the-loop control and efficient skill acquisition. To bridge this gap\, we explore human-to-robot motion mapping. Through user studies\, we demonstrate that a kinematic-twin interface\, which itself leverages the hand’s modularity\, significantly reduces cognitive workload\, and improves task success compared to vision-based motion retargeting. Using kinematic-twin interfaces\, I will show how various dexterous skills can be acquired through learning from human demonstrations. Finally\, I discuss the integration of multimodal tactile sensing into the DeltaHands framework to enable fine-grained manipulation. To exploit the transient yet information-rich nature of touch\, I present an object representation learning method which extracts and preserves meaningful tactile signals during exploratory interactions to inform and guide downstream\, object-centric manipulation policies. In conclusion\, the concise and modular design of the non-anthropomorphic robotic hands\, sensing systems\, and control interfaces enables the direct transfer of human dexterity to robotic platforms and supports dexterous manipulation learning. \n\nCommittee: \nZeynep Temel (co-chair)\, Carnegie Mellon University \nOliver Kroemer (co-chair)\, Carnegie Mellon University \nNancy Pollard\, Carnegie Mellon University \nOliver Brock\, Technische Universität Berlin \nThesis Draft
URL:https://www.ri.cmu.edu/event/designing-a-learning-enabled-non-anthropomorphic-robotic-hand-framework-for-dexterous-manipulation/
LOCATION:Gates-Hillman Center 8102
CATEGORIES:PhD Thesis Defense,Student Talks
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=America/New_York:20260518T100000
DTEND;TZID=America/New_York:20260518T113000
DTSTAMP:20260707T165816
CREATED:20260508T181152Z
LAST-MODIFIED:20260508T181152Z
UID:151234-1779098400-1779103800@www.ri.cmu.edu
SUMMARY:Deep Abstraction Learning for Neuro-Symbolic World Modeling
DESCRIPTION:Abstract: Modern foundation models have achieved remarkable progress by learning broad physical and semantic common sense from large-scale data. However\, robots operating in open-ended environments require more than general knowledge alone: they must continually specialize in new tasks\, environments\, and experiences encountered during deployment. Given only limited deployment-time data\, how can robots learn to solve substantially more—and conceptually harder—problems than those seen during training? \nThis thesis addresses this challenge through deep abstraction learning\, where robots discover relational abstractions grounded by deep neural networks to construct neuro-symbolic world models from high-dimensional and noisy observations. By ignoring task-irrelevant details\, abstractions can be learned efficiently from limited experience while enabling abstract planning for long-horizon decision-making problems involving many interacting objects. The central hypothesis is that abstractions and world models should co-evolve: abstractions enable efficient reasoning and planning\, while planning structure and execution failures drive the discovery of richer abstractions. \nThis framework consists of three key components. Deep state abstractions\, represented as relational predicates\, map high-dimensional observations into symbolic concepts that support reasoning and planning under noisy sensory inputs. Deep action abstractions\, represented as relational option policies\, capture reusable behaviors that enable robots to recover from failures and progressively acquire new abstractions from interaction. Neuro-symbolic world models describe how action abstractions transform state abstractions\, enabling abstract planning that generalizes to unseen long-horizon tasks. To study these challenges\, this thesis introduces benchmark suites that reveal the limitations of purely neural approaches and motivate abstraction-based world modeling. \nBuilding on these foundations\, this dissertation proposes two future directions. The first studies how pre-trained coding agents can synthesize expressive programmatic world models that support planning with recomposable tools such as abstractions and perception models. The second\, RoboSymphony\, studies multi-agent decision-making in which robots learn abstractions over the intentions and behaviors of other agents and humans\, enabling coordination in collaborative tasks. \nTogether\, these contributions advance a unified neuro-symbolic approach for robots that learn and plan with deep abstractions\, enabling efficient specialization and decision-making in open-ended real-world environments. \nThesis Committee: \nSebastian Scherer (Chair)\nMaxim Likhachev\nKatia Sycara\nTom Silver\, Princeton University\nLeslie P. Kaelbling\, Massachusetts Institute of Technology\n\nThesis Draft
URL:https://www.ri.cmu.edu/event/deep-abstraction-learning-for-neuro-symbolic-world-modeling/
LOCATION:Newell-Simon Hall 4305
CATEGORIES:PhD Thesis Proposal,Student Talks
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=America/New_York:20260515T120000
DTEND;TZID=America/New_York:20260515T130000
DTSTAMP:20260707T165816
CREATED:20260506T144730Z
LAST-MODIFIED:20260506T144730Z
UID:151206-1778846400-1778850000@www.ri.cmu.edu
SUMMARY:Robot Learning and Wearable Interfaces in Pursuit of Robotic Caregivers
DESCRIPTION:Abstract:  Designing safe and reliable robotic assistance for caregiving is a grand challenge in robotics. A sixth of the United States population is over the age of 65 and more than 1 in 4 (over 70 million) adults in the United States reported having a disability in 2022. Robotic caregivers could positively benefit society; yet\, physical robotic assistance presents several challenges and open research questions relating to autonomous control\, multimodal sensing and learning\, and accessible interfaces. In this talk\, I will present recent techniques and technology that my group has developed towards addressing core challenges in robotic caregiving. First\, I will introduce inertial and high-density electromyography (HDEMG) wearable interfaces that enable people with severe loss of motor and hand function (due to spinal cord injury or neurodegenerative diseases) to embody physically assistive mobile manipulators in their home. I will then present our recent work in robot learning\, including online and offline policy learning\, to perform complex manipulation in assistive scenarios. This includes learning reward functions and robot control policies\, sim-to-real transfer\, a new technique to scale imitation learning for bimanual manipulation\, and new opportunities presented by generative simulation. \n  \nHomepage:  http://zackory.com
URL:https://www.ri.cmu.edu/event/robot-learning-and-wearable-interfaces-in-pursuit-of-robotic-caregivers/
LOCATION:Newell-Simon Hall 4305
CATEGORIES:Faculty Events
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=America/New_York:20260513T153000
DTEND;TZID=America/New_York:20260513T163000
DTSTAMP:20260707T165816
CREATED:20260429T191158Z
LAST-MODIFIED:20260429T191629Z
UID:151146-1778686200-1778689800@www.ri.cmu.edu
SUMMARY:Quanta Perception as Probabilistic Events
DESCRIPTION:Abstract:  Autonomous systems ultimately rely on extracting information from light\, yet remain brittle in extreme environments\, from nighttime navigation to high-speed robotics. This limitation stems from a classical imaging abstraction: conventional sensors integrate photon flux over fixed exposure windows\, imposing trade-offs between sensitivity\, dynamic range\, and temporal resolution that degrade perception when photons are scarce or dynamics are rapid. Emerging quanta (single-photon) image sensors overcome these limits by detecting individual photons\, but they generate photon streams that exceed the compute and latency budgets of real-time systems by orders of magnitude. \n\nHere we introduce probabilistic events\, a computational primitive for real-time quanta perception at the limit of individual photons. By computing the posterior distribution over the time since the last abrupt intensity change\, we represent photon streams as recursively computed belief states. Rather than the binary\, fixed-threshold triggers of event cameras\, this recursive Bayesian formulation yields three simultaneous\, low-latency signals: motion-adaptive scene flux\, high-fidelity activity maps\, and an entropy measure quantifying perceptual uncertainty. This representation enables perception in extreme conditions\, including detecting and estimating the pose of a running person at ~0.05 lux illumination—without retraining standard vision models. Our approach sustains input throughputs exceeding 50\,000 quanta frames per second on commodity GPU hardware—four to five orders of magnitude faster than state-of-the-art quanta reconstruction baselines—yielding kilohertz-scale outputs even for megapixel arrays. By replacing frame reconstruction with direct probabilistic inference over photon streams\, this work enables real-time perception at the photon limit and bridges photon-counting quanta sensing with practical robotic vision.\n \nBio:   Varun Sundar is a graduate student at the University of Wisconsin–Madison\, pursuing a Ph.D. in computer science. At UW–Madison\, he is advised by Prof. Mohit Gupta\, where he focuses on single-photon imaging techniques. His work has been published at venues such as CVPR\, ICCV\, and SIGGRAPH\, and has included live demos at ICCP 2023\, CVPR 2024 and SIGGRAPH 2024 (which won the best-in-show award in the Emerging Technologies track). In 2026\, he was awarded the Ivanisevic Award at UW–Madison\, which recognizes outstanding computer science dissertators. He previously received a bachelor’s degree in electrical engineering from the Indian Institute of Technology\, Madras in 2020. \nHomepage:   https://varun19299.github.io/ \nSponsor:\nThe VASC seminar is generously sponsored by HeyGen\, an all-in-one AI-powered video generation platform that leverages advances in computer vision\, generative modeling\, and multimodal learning to make high-quality video creation both scalable and accessible.
URL:https://www.ri.cmu.edu/event/quanta-perception-as-probabilistic-events/
LOCATION:3305 Newell-Simon Hall
CATEGORIES:Seminar,VASC Seminar
ATTACH;FMTTYPE=image/jpeg:https://www.ri.cmu.edu/app/uploads/2026/04/5-13-26.jpg
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=America/New_York:20260513T090000
DTEND;TZID=America/New_York:20260513T223000
DTSTAMP:20260707T165816
CREATED:20260508T135524Z
LAST-MODIFIED:20260508T135524Z
UID:151228-1778662800-1778711400@www.ri.cmu.edu
SUMMARY:Learning Bayesian Experimental Design Policies Efficiently
DESCRIPTION:Abstract:\nBayesian Experimental Design (BED) provides a principled framework for informative data collection\, and is applied across domains as varied as adaptive clinical trials\, ecological monitoring\, hyperparameter optimization\, and robotic search. Despite this broad applicability\, BED methods remain difficult to deploy in practice: high-quality decision-making is computationally expensive\, calibrated uncertainty estimation in sequential decision problems is challenging\, and classical BED methods do not have the mechanisms to leverage the rich\, unstructured prior knowledge available in real-world problems. This thesis develops methods that make BED more efficient along each of these axes. \nWe first present a case study in multi-robot active search\, casting the problem as a decentralized batch BED task. Our proposed algorithm\, based on myopic posterior sampling\, performs robustly under communication and hardware failures\, validated through field tests with heterogeneous teams of UGVs and UAVs in a 75\,000 m² forested environment. This case study makes a recurring tension in BED concrete: practitioners are routinely forced to trade off decision-making quality to stay within real-time computational budgets. To close this gap\, we learn amortized BED policies by leveraging the domain-permutation equivariance inherent in BED problems\, which yields markedly more sample-efficient policy and Q-value architectures. We field-test these amortized policies\, achieving wall-clock speed-ups of up to 2.86× over our previous baselines to reach the same recall performance. \nBeyond active search\, we make two further contributions. First\, we introduce Epistemic Bellman equations: a model-based framework for Q-value uncertainty quantification that produces well-calibrated estimates in the bootstrapped Q-learning setting. Our estimates make value uncertainty usable as a primitive in downstream tasks such as robust policy optimization and efficient exploration. Second\, we develop methods for incorporating large-language-model-elicited priors into Bayesian Optimization\, leveraging the rich auxiliary information — expert knowledge\, scientific literature\, intermediate diagnostics\, training curves — that is abundant in real-world optimization problems but discarded by standard BO algorithms. Across HPOBench and a real-world nuclear-fusion tokamak stabilization task\, our methods consistently outperform both standard BO and prior LLM-based BO approaches. \nTogether\, these contributions broaden the scope of the BED toolkit in the real world. \nCommittee:\nJeff Schneider (Chair)\, Carnegie Mellon University\nDavid Held\, Carnegie Mellon University\nAviral Kumar\, Carnegie Mellon University \nMatthias Seeger\, Amazon \nThesis Draft
URL:https://www.ri.cmu.edu/event/learning-bayesian-experimental-design-policies-efficiently/
LOCATION:3305 Newell-Simon Hall
CATEGORIES:PhD Thesis Defense,Student Talks
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=America/New_York:20260512T130000
DTEND;TZID=America/New_York:20260512T140000
DTSTAMP:20260707T165816
CREATED:20260507T180227Z
LAST-MODIFIED:20260507T180227Z
UID:151217-1778590800-1778594400@www.ri.cmu.edu
SUMMARY:Diffusion Temperature Sampling and Projective Ray Positional Encoding for Multi-view Attention
DESCRIPTION:Abstract: Inductive biases have proven effective and often essential in the design of performant deep learning systems. This thesis presents two contributions that target distinct facets of how inductive biases can improve modern deep models. In Part 1\, we focus on controlling the diversity–likelihood trade-off at inference-time of generative models. In Part 2\, we study the architectural inductive bias for 3D perception for multi-view transformer models.\nPart 1 presents Temporal Score Rescaling (TSR)\, a mechanism to steer the sampling diversity of denoising diffusion and flow matching models\, allowing users to sample from a sharper or broader distribution than the training distribution. We build on the observation that these models leverage (learned) score functions of noisy data distributions for sampling\, and show that rescaling these allows one to effectively control a `local‘ sampling temperature. Notably\, this approach does not require any finetuning or alterations to training strategy\, and can be applied to any off-the-shelf model. We validate our framework across five disparate tasks — image generation\, pose estimation\, depth prediction\, robot manipulation\, and protein design. Across these tasks\, our approach allows sampling from sharper (or flatter) distributions and yields consistent performance gains.\nPart 2 presents RayRoPE\, where we study positional encodings for multi-view transformers that process tokens from a set of posed input images\, and seek a mechanism that encodes patches uniquely\, allows SE(3)-invariant attention with multi-frequency similarity\, and can adapt to the geometry of the underlying 3D scene. We find that prior encoding schemes for multi-view attention do not meet these desiderata. RayRoPE represents patch positions based on associated rays and computes query-frame projective coordinates to ensure $SE(3)$ invariance. To adapt to scene geometry\, RayRoPE predicts a per-token depth to obtain its position along the corresponding ray\, while also modeling uncertainty and analytically computing the expected positional encoding. We validate our method on the tasks of novel-view synthesis and stereo depth estimation. While remaining efficient\, RayRoPE consistently improves over alternate position encoding schemes. \n\nCommittee:\nProf. Shubham Tulsiani (chair)\nProf. Deva Ramanan\nProf. David Held\nHanzhe Hu
URL:https://www.ri.cmu.edu/event/diffusion-temperature-sampling-and-projective-ray-positional-encoding-for-multi-view-attention/
LOCATION:Newell-Simon Hall 4305
CATEGORIES:MSR Thesis Presentation,Student Talks
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=America/New_York:20260511T100000
DTEND;TZID=America/New_York:20260511T113000
DTSTAMP:20260707T165816
CREATED:20260409T202454Z
LAST-MODIFIED:20260504T190942Z
UID:150973-1778493600-1778499000@www.ri.cmu.edu
SUMMARY:Leveraging Tactile Sensing to Resolve Uncertainty in Contact-Rich Manipulation
DESCRIPTION:Abstract: Manipulation in agricultural and unstructured environments often involves contact-rich interactions with occluded objects. Most deployed systems treat contact as a hazard and rely on vision alone\, which limits deployment from real-world field settings. This thesis adopts a different perspective: contact is a strategy for obtaining information where vision cannot provide. Touch can reveal additional spatial\, temporal\, and semantic cues\, along with other task relevant latent states. It argues that augmenting vision with contact sensing\, and conditioning learned policies on structured representations of task-relevant hidden states (inferred from sequential contacts)\, enables reliable manipulation where vision alone cannot disambiguate hidden states. In POMDP settings\, we show that structured representations of task-relevant hidden states outperform end-to-end approach conditioned on raw observation history.\n\nThe dissertation develops this argument in three movements: establishing the limits of vision-only manipulation through a precise agricultural task where occlusion makes sub-centimeter alignment unreliable; showing that vibrotactile sensing via contact microphone arrays recovers both the spatial and semantic cues; and demonstrating that conditioning policies on structured hidden-state representations yields history-aware adaptive behavior on long-horizon tasks that end-to-end policies cannot reliably solve.\n\n\n\n\nCommittee:\n\n\nOliver Kroemer (chair)\, Carnegie Mellon University\nGeorge Kantor (co-chair)\, Carnegie Mellon University\nYonatan Bisk\, Carnegie Mellon University\nTapomayukh Bhattacharjee\, Cornell University \n\n\nThesis Draft
URL:https://www.ri.cmu.edu/event/ri-phd-thesis-defense-mark-lee/
LOCATION:NSH 3305
CATEGORIES:PhD Thesis Defense,Student Talks
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=America/New_York:20260511T083000
DTEND;TZID=America/New_York:20260511T100000
DTSTAMP:20260707T165816
CREATED:20260409T174428Z
LAST-MODIFIED:20260503T203935Z
UID:150953-1778488200-1778493600@www.ri.cmu.edu
SUMMARY:Think Globally\, Solve Locally: Non-sequential Planning for Robotic Manipulation
DESCRIPTION:Abstract:\n\nRobotic manipulation requires reasoning that bridges local competence over fine-grained dynamics with the  construction of valid long-horizon plans. This mirrors human reasoning\, where fast\, automatic processes propose locally plausible actions from experience\, while slower deliberation integrates them into a whole. Furthermore\, evidence from cognitive science suggests that humans do not reason sequentially from start to finish; instead\, they anchor on intermediate landmarks\, plan from multiple directions\, and let local insights reshape global structure.\n\nYet\, current approaches in robotics typically formulate planning as a monolithic\, unidirectional search from an initial state toward a goal or in the opposite direction. This thesis argues for a departure from that paradigm. We propose Non-Sequential Deliberative Planning\, a framework that distributes deliberation across multiple local regions in the problem space\, rather than committing to a systematic\, directed reasoning. By exploring simultaneously from these regions\, local solvers (whether geometric\, heuristic\, or generative) operate where they are most effective\, while a global search composes them into a coherent plan with formal guarantees. \nWe instantiate this principle across three algorithmic regimes. For high-dimensional motion planning\, we introduce Multi-Graph Search (MGS)\, which identifies key states as intermediate landmarks and simultaneously grows search trees from each\, merging local subgraphs into a global solution with provable completeness and bounded suboptimality guarantees.\nSecond\, for contact-rich manipulation\, we present MOSAIC\, which treats physics-validated skills as local competences. It composes sequences of skills\, such as pushing or grasping\, through a non-sequential search that connects local regions of reliable execution. Third\, for scenarios where deliberation time is severely limited\, we develop methodologies that shift non-sequential reasoning to an offline phase. By integrating manipulation behaviors directly into preprocessing\, we generate motions whose manipulation outcomes are provably reliable\, and the deliberation time is guaranteed to be within a user-defined time bound.\nTo complete this thesis\, we consider three extensions. First\, MOSAIC relies on physics simulation to estimate the outcome of contact-rich interactions during online planning—a significant computational bottleneck. We propose to address this by utilizing an offline phase to learn proxies and construct data structures that enable efficient online planning over long horizons. Second\, manipulation skills are typically designed and learned for interactions between a robot and a single object. Real-world deployment\, however\, brings scenes with many movable objects\, and tracking them jointly causes a combinatorial explosion in the planning state space. To address this\, we propose to extend our prior framework with partial state planning\, in which the global search operates over decoupled\, object-centric representations and evaluates multi-object interactions only when necessary. Third\, drawing on our prior work in multi-robot coordination—Experience-Accelerated Multi-Robot Planning (xECBS) and Multi-Robot Multi-Model Diffusion (MMD)—we plan to extend the non-sequential planning architecture to multi-arm settings\, enabling concurrent execution across multiple manipulators. \nUltimately\, this thesis establishes a unified framework for non-sequential decision making\, composing fast local competences into globally sound manipulation plans across a range of real-world settings. \n\n\nThesis Committee:\n\n\nProf. Maxim Likhachev (Chair)\nProf. Changliu Liu\nProf. David Held\nProf. Oren Salzman (Technion University)\n\n\n\nThesis proposal document draft
URL:https://www.ri.cmu.edu/event/ri-phd-thesis-proposal-itamar-mishani/
LOCATION:Newell-Simon Hall 1305
CATEGORIES:PhD Thesis Proposal,Student Talks
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=America/New_York:20260504T163000
DTEND;TZID=America/New_York:20260504T180000
DTSTAMP:20260707T165816
CREATED:20260427T175100Z
LAST-MODIFIED:20260427T175100Z
UID:151125-1777912200-1777917600@www.ri.cmu.edu
SUMMARY:Forecasting at Scale with Efficient Deep Learning Architectures
DESCRIPTION:Abstract:\nTime Series Foundation Models (TSFMs) have scaled rapidly\, with publicly reported pretraining corpora growing from 1.23 billion to 1 trillion data points between 2024 and 2026\, an approximately 800× increase in two years. Recent work has further supplemented real-world data with synthetic data to expose models to broader time series patterns. Yet\, this data-centric paradigm raises a fundamental question: must intelligent forecasting rely solely on scale\, or can intentional architectural design unlock better generalization? This thesis proposes that more intelligently and efficiently leveraging existing data\, rather than scale alone\, is key to achieving better forecasting generalization. We pursue this through three parallel architectural themes: exploiting cross-channel structure beyond temporal patterns\, enabling zero-shot generalization through structured composition\, and reducing gradient and forecast variance by design. Each theme aims to enhance generalization with available data while treating computational efficiency as a core design principle. \nIn this thesis\, we demonstrate that scale is not the only path to generalization by: developing multivariate architectures that leverage cross-channel dependencies efficiently while reducing forecast error; showing that architectures can generalize beyond their training distribution in both patterns and concepts; and verifying variance-aware architectural designs that extract richer training signals from existing data\, provably reducing gradient variance while reducing forecast error and improving calibration. \nWithin the first theme\, we further propose pretraining strategies for multivariate TSFMs to investigate whether data balancing and curriculum learning can improve downstream generalization given the same pretraining corpora. Within the second theme\, we propose an additional dimension of generalization\, extending beyond pattern and concept generalization to horizon generalization\, an important consideration for TSFMs applied across diverse tasks and domains. Overall\, this work contributes new insights into advancing time series forecasting generalization through efficient architectural design. \n\n\nCommittee:\nArtur Dubrawski\, Chair\nJohn Dolan\nBarnabás Póczos\nMichael W. Mahoney (University of California\, Berkeley)\n\n\nThesis Link
URL:https://www.ri.cmu.edu/event/forecasting-at-scale-with-efficient-deep-learning-architectures/
LOCATION:GHC 4405
CATEGORIES:PhD Thesis Proposal,Student Talks
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=America/New_York:20260504T133000
DTEND;TZID=America/New_York:20260504T150000
DTSTAMP:20260707T165816
CREATED:20260427T201525Z
LAST-MODIFIED:20260427T201525Z
UID:151127-1777901400-1777906800@www.ri.cmu.edu
SUMMARY:Leveraging Local Models for Planning and Control with Contact
DESCRIPTION:Abstract: Many planning and control approaches in robotics have converged on optimization-based formulations\, with recent advances achieved by leveraging significant data and compute to attempt to tackle these nonlinear and non-convex problems. In this thesis\, we instead focus on local models and demonstrate their benefits and surprising effectiveness. In the case of smooth optimization\, the local model is a convex quadratic program. We show how this structure enables efficient parallelization and scaling by mapping it to a neural network\, and that a fixed linear model is still capable of basic locomotion tasks even with a large sim-to-real gap. We then look at the non-smooth case that arises in contact-implicit approaches which can be expressed as quadratic programs with complementarity constraints and develop a C++ solver\, Marble\, that leverages the structure of relaxed complementarity. Finally\, we propose future work that unifies existing hard and soft contact models under a common framework and examines them in the context of planning. We also propose applying the resulting Marble solver for local motion retargeting tasks\, exploring applications in both simulation and on hardware in an iterative learning control context. \nThesis Committee: \nZac Manchester (chair)\nAaron Johnson\nLorenz Biegler\nPat Wensing (University of Notre Dame)\n\nThesis URL
URL:https://www.ri.cmu.edu/event/leveraging-local-models-for-planning-and-control-with-contact/
LOCATION:3305 Newell-Simon Hall
CATEGORIES:PhD Thesis Proposal,Student Talks
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=America/New_York:20260504T090000
DTEND;TZID=America/New_York:20260504T103000
DTSTAMP:20260707T165816
CREATED:20260430T162020Z
LAST-MODIFIED:20260430T162020Z
UID:151157-1777885200-1777890600@www.ri.cmu.edu
SUMMARY:WinkTPG: An Execution Framework for Multi-Agent Path Finding Using Temporal Reasoning
DESCRIPTION:Abstract: \nPlanning collision-free paths for a large group of agents is a challenging problem in many real-world applications. While recent advances in Multi-Agent Path Finding (MAPF) have shown promising progress\, standard MAPF planners continue to rely on simplified kinodynamic models\, preventing agents from directly following the generated MAPF plan. To bridge this gap\, we propose kinodynamic Temporal Plan Graph planning (kTPG)\, a multi-agent speed optimization algorithm that efficiently refines a MAPF plan into a set of kinodynamically feasible speed profiles. We further incorporate execution timing uncertainty models and provide deterministic guarantees under bounded uncertainty models and probabilistic guarantees under stochastic models. Building on kTPG\, we propose Windowed kTPG (WinkTPG)\, a MAPF execution framework that incrementally refines MAPF plans using a window-based mechanism\, dynamically incorporating agent information during execution to reduce uncertainty. Experiments show that WinkTPG can generate speed profiles for up to 1\,000 agents within 1 second and improve solution quality by up to 51.7% over existing MAPF execution methods. \nCommittee: \nProf. Jiaoyang Li (co-chair) \nProf. Stephen F. Smith (co-chair) \nProf. John Dolan \nHsu-Kuang Chiu
URL:https://www.ri.cmu.edu/event/winktpg-an-execution-framework-for-multi-agent-path-finding-using-temporal-reasoning/
LOCATION:Newell-Simon Hall 4305
CATEGORIES:PhD Speaking Qualifier,Student Talks
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=America/New_York:20260501T120000
DTEND;TZID=America/New_York:20260501T130000
DTSTAMP:20260707T165816
CREATED:20260420T194659Z
LAST-MODIFIED:20260420T194659Z
UID:151089-1777636800-1777640400@www.ri.cmu.edu
SUMMARY:Precise and Generalizable Robot Manipulation
DESCRIPTION:Abstract:   Robots in factories are still largely limited to structured environments with known object models. How can we bring robots into the more diverse\, unstructured settings of our daily lives\, where objects may vary widely in shape and appearance\, while maintaining reliable performance? A popular direction today is to train generalist robot policies on large-scale internet data and broad robot datasets. However\, today’s generalist policies still lack the precision needed for robust real-world operation. In this talk\, I argue that closing this gap requires learning a hierarchy over robot motion: learning both what subgoals to achieve as well as how to move the robot end-effector to achieve them. I will present hierarchical motion policies that combine high-level subgoal prediction with a learned low-level policy. I will show how this hierarchical approach has enabled us to achieve both generalizable and precise object manipulation.
URL:https://www.ri.cmu.edu/event/precise-and-generalizable-robot-manipulation/
LOCATION:Newell-Simon Hall 4305
CATEGORIES:Faculty Events
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=America/New_York:20260430T160000
DTEND;TZID=America/New_York:20260430T173000
DTSTAMP:20260707T165816
CREATED:20260421T193956Z
LAST-MODIFIED:20260422T143815Z
UID:151107-1777564800-1777570200@www.ri.cmu.edu
SUMMARY:Longitudinal Human–Robot Interaction: Adaptive Personalization Across Repeated Encounters
DESCRIPTION:Abstract: \nAs robots increasingly move into homes\, healthcare settings\, and public environments\, many are expected to support people not through single encounters\, but through repeated interaction over time. In these settings\, successful human–robot interaction depends not only on immediate task performance\, but also on how users adapt to robotic systems\, how expectations change with repeated exposure\, and how interaction preferences evolve across sessions and contexts. Despite growing interest in personalization\, many robotic systems still assume that user preferences can be estimated once and treated as stable\, with limited understanding of how preferences develop longitudinally.\n\nThis thesis investigates longitudinal human–robot interaction by examining how user preferences\, engagement\, and interaction strategies change through repeated encounters with socially interactive robots. In robotic exercise support for older adults\, an exploratory Wizard-of-Oz study revealed substantial variation in how participants naturally engaged with a conversational exercise robot\, ranging from brief task-focused exchanges to extended social interaction. Building on these findings\, a four-week longitudinal study compared two contrasting robot personalities during repeated exercise sessions: a Social Buddy Personality emphasizing companionship and conversation\, and an Exercise Coach Personality emphasizing structured feedback and task-focused guidance. Participants responded positively to both personalities but valued different aspects of each\, with preferences varying across individuals and shifting across sessions.\n\nTo examine whether similar temporal dynamics extend beyond exercise\, this thesis also investigates repeated interaction in accessibility robotics through a longitudinal navigation study with blind and low-vision users. Across multiple weeks of navigation in public environments\, participants demonstrated evolving preferences for delegation and autonomy\, with assistance strategies changing according to context\, familiarity\, and accumulated experience. Together\, these studies show that user preferences in sustained human–robot interaction are not static\, but develop through repeated exposure and situational adaptation.\n\nMotivated by these findings\, this thesis proposes a multi-timescale adaptive robotic exercise coaching framework that dynamically adjusts social interaction and coaching behavior using multimodal estimates of user state\, behavior\, and engagement. By integrating exercise performance\, conversational behavior\, and interaction history\, the proposed system models preference across long-term trends\, session-level variation\, and moment-to-moment interaction. Overall\, this work contributes new insight into how temporally aware personalization can support long-term human–robot interaction in health\, accessibility\, and everyday assistive contexts.\n\n\nThesis Committee:\nAaron Steinfeld\, chair \nReid Simmons\, CMU\nHenny Admoni\, CMU\nMaja Matarić\, USC\nTaskin Padir\, Amazon Robotics\, Northeastern University\n\n\nThe link to the document can be found here: thesis_proposal 
URL:https://www.ri.cmu.edu/event/longitudinal-human-robot-interaction-adaptive-personalization-across-repeated-encounters/
LOCATION:1305 Newell Simon Hall
CATEGORIES:PhD Thesis Proposal,Student Talks
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=America/New_York:20260430T140000
DTEND;TZID=America/New_York:20260430T160000
DTSTAMP:20260707T165816
CREATED:20260429T011037Z
LAST-MODIFIED:20260429T011037Z
UID:151137-1777557600-1777564800@www.ri.cmu.edu
SUMMARY:Robotics Major and Additional Major Capstone Presentations
DESCRIPTION:Undergraduates in the Robotics Major and Robotics Additional Major will be demonstrating their capstone projects. Our majors have designed and built robots for their chosen tasks. We have a frisbee throwing robot\, water sampling boat\, brachiating mechanism\, haptic-feedback arm and hand\, an autonomous backpack\, and a drover (drone + rover pair). \nDemonstrations will be in the NEW Robotics Education Lab on the first floor in NSH 1201. Come see robotics\, congratulate our students\, and check out the new undergraduate lab space!
URL:https://www.ri.cmu.edu/event/robotics-major-and-additional-major-capstone-presentations/
LOCATION:NSH 1201
CATEGORIES:Special Events
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=America/New_York:20260429T140000
DTEND;TZID=America/New_York:20260429T153000
DTSTAMP:20260707T165816
CREATED:20260410T190951Z
LAST-MODIFIED:20260421T200628Z
UID:150986-1777471200-1777476600@www.ri.cmu.edu
SUMMARY:Leveraging Deformations in Soft Objects and Robots for Contact-Rich Dexterity
DESCRIPTION:Abstract:\nRobust dexterity depends on the ability to cause and reason about deformation. Robots may routinely need to interact with cables\, dough\, textiles\, and hair in our homes\, where deformation is the object state. Soft robots\, in turn\, are built from compliant materials in which deformation is not a by-product of interaction but its primary mechanism. In this talk\, I will discuss mechanics-aware approaches to contact-rich dexterous manipulation with soft objects and soft robots\, using mechanics-based priors as a physically grounded intermediate layer between raw sensing and task-level control. These priors take the form of mesh structures with geometric regularization\, mechanics modeling\, and volumetric dynamics models learned from simulation. I will show how these priors enable robots to perceive deformation through occlusion\, predict it\, and transfer skills across soft objects\, soft robot bodies\, and human demonstrations. I will conclude by discussing future directions for mechanics-aware models and robot skill learning frameworks that may scale to diverse deformable objects and compliant morphologies.\n\n\n\nCommittee:\n\nJean Oh (co-chair)\, Carnegie Mellon University\nJeffrey Ichnowski (co-chair)\, Carnegie Mellon University\nChristopher G. Atkeson\, Carnegie Mellon University\nYunzhu Li\, Columbia University\n\n\n\nThesis Draft
URL:https://www.ri.cmu.edu/event/ri-phd-thesis-defense-uksang-yoo/
LOCATION:GHC 6121
CATEGORIES:PhD Thesis Defense,Student Talks
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=America/New_York:20260429T120000
DTEND;TZID=America/New_York:20260429T133000
DTSTAMP:20260707T165816
CREATED:20260319T161254Z
LAST-MODIFIED:20260427T150832Z
UID:150637-1777464000-1777469400@www.ri.cmu.edu
SUMMARY:Integrating Reinforcement Learning and Model Predictive Control for Autonomous Off-road Driving
DESCRIPTION:Abstract: Safe and effective autonomous traversal of off-road terrain is challenging due to both terrain properties\, such as low traction in sand or deformability of mud\, and terrain geometries\, including steep slopes\, ditches\, and uneven surfaces that can induce unsafe vehicle behaviors like excessive pitch and roll. Model Predictive Path Integral (MPPI) control provides a powerful framework for solving Model Predictive Control (MPC) problems and has demonstrated strong performance in off-road and agile locomotion tasks. Key to MPPI’s success are the parallelizable open-loop dynamics rollouts\, the optimization costs for which are generally determined from a learned or predefined cost map of the terrain. While MPPI can be implemented in a purely physics-based formulation\, there is growing interest in integrating data-driven methods to improve performance and adaptability.\nIn this talk\, we present Value Function-Guided MPPI\, a hierarchical framework that integrates reinforcement learning (RL) with MPPI. Rather than relying on an explicit cost map\, the planner uses a learned RL value function as an implicit\, execution-aware objective\, allowing trajectories to be evaluated based on the expected performance of the controller. MPPI performs trajectory exploration\, while a pretrained RL policy executes the selected plans\, creating a bidirectional feedback loop between planning and control. This removes the need for manual cost design and better aligns planning with execution dynamics. We evaluate the method in simulation and real-world field tests on a full-size Yamaha ATV\, showing improved safety and goal-reaching performance in challenging terrain\, and discuss sim-to-real challenges encountered. \nFinally\, we outline ongoing and future work that draws upon a growing body of literature reframing MPC itself as a learning problem\, using RL to optimize parameters of the MPC formulation. Across the approaches discussed in this talk\, we highlight key tradeoffs and design choices when integrating RL and MPC\, such as sim-to-real transfer\, robustness\, and adaptability\, thereby informing how algorithms should allocate responsibility between learning and control in real-world autonomous systems. \n \nCommittee:\nJeff Schneider (co-chair)\nGuanya Shi (co-chair)\nWenshan Wang\nSam Triest
URL:https://www.ri.cmu.edu/event/ri-phd-speaking-qualifier-anoushka-avavilli/
LOCATION:Gates Hillman Center 6115
CATEGORIES:PhD Speaking Qualifier,Student Talks
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=America/New_York:20260428T150000
DTEND;TZID=America/New_York:20260428T160000
DTSTAMP:20260707T165816
CREATED:20260423T163219Z
LAST-MODIFIED:20260423T163219Z
UID:151113-1777388400-1777392000@www.ri.cmu.edu
SUMMARY:Efficient Warping for Visual Perception
DESCRIPTION:Abstract: \nSalient foreground regions (e.g.\, vehicles\, faces) occupy only ~10% of pixels\, while less informative backgrounds (e.g.\, sky\, trees) dominate ~90%. This imbalance fundamentally limits visual perception: \n(1) 2D discriminative tasks (e.g.\, domain-adaptive detection and segmentation) rely heavily on large background regions with high cross-domain variation\, making domain adaptation difficult. \n(2) 2D generative tasks (e.g.\, image-to-image translation) compress images into a latent space where small foreground regions receive even fewer spatial resources\, making it difficult to preserve and reconstruct fine-grained details. \n(3) 3D tasks (e.g.\, occupancy prediction) allocate substantial computation on less informative background regions\, leading to significant latency and memory overhead. \nIn this work\, we propose an efficient image warping framework with instance-level saliency to oversample salient foregrounds and undersample less informative backgrounds. Our method is model-agnostic\, works with arbitrary saliency priors\, requires no architecture modification\, and introduces negligible computational\, memory\, and latency overhead. Experiments on domain-adaptive object detection and semantic segmentation\, image-to-image translation (e.g.\, human and driving scene relighting\, and driving scene translation)\, and 3D occupancy prediction demonstrate improved accuracy for discriminative and 3D tasks\, and enhanced fidelity and realism for generative tasks\, while maintaining high efficiency. \nCommittee: \nProf. Srinivasa Narasimhan \nProf. Deva Ramanan \nProf. Shubham Tulsiani \nGaurav Parmar
URL:https://www.ri.cmu.edu/event/efficient-warping-for-visual-perception/
LOCATION:Newell-Simon Hall 4305
CATEGORIES:PhD Speaking Qualifier,Student Talks
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=America/New_York:20260428T120000
DTEND;TZID=America/New_York:20260428T133000
DTSTAMP:20260707T165816
CREATED:20260325T185125Z
LAST-MODIFIED:20260421T154725Z
UID:150738-1777377600-1777383000@www.ri.cmu.edu
SUMMARY:Evolutionary Environment Optimization for Large-Scale Multi-Robot Systems
DESCRIPTION:Abstract:\n\n\nRecent advances in robotics have enabled researchers and practitioners to deploy large-scale multi-robot systems\, where hundreds to thousands of robots operate simultaneously in a shared environment. These systems are increasingly used in applications that require high efficiency and reliability\, such as automated warehouses\, robotic sorting systems\, and autonomous transportation. A fundamental challenge in such systems is to enable many robots to move efficiently while sharing limited space\, avoiding congestion\, and continually finishing new tasks. \nWhile the community has spent a significant amount of attention on studying how to more efficiently and effectively coordinate robots in multi-robot systems\, very few works focus on the environment where the systems are deployed. By environment\, we refer to both physical elements\, such as spatial arrangement of storage shelves in automated warehouses\, and virtual elements\, such as traffic rules that influence how robots move and interact. A well-designed environment can significantly reduce congestion\, foster implicit coordination among robots\, and improve system throughput. In this thesis\, we study the problem of Environment Optimization for large-scale multi-robot systems. We identify three main components of the environment that we find optimizable. We formally define each of them as black-box optimization problems and apply evolutionary-based optimizers to solve them. \nWe first discuss Task Mapping Optimization in the context of robotic sorting systems (RSS)\, where robots transport packages from induct workstations to eject chutes according to shipping destinations. The destination-to-chute assignment directly shapes traffic demand: poor assignments overload certain regions\, create conflicts\, and reduce throughput. To fill this gap\, we formally define the problem of Task Mapping Optimization (TMO) and propose methods to search for high-quality task mappings that improve throughput by balancing robot traffic. \nWe then discuss Layout Optimization in the context of automated warehouses\, where robots transport packages or inventory pods between locations. In these systems\, layout strongly affects robot coordination: narrow passages create bottlenecks and poorly placed shelves induce congestion\, even when state-of-the-art planning algorithms are applied. Existing warehouse layouts often follow grid-like structures designed for human accessibility\, but such patterns are not necessarily suitable for robots\, which do not share human ergonomic constraints. This motivates the question of how a warehouse should be designed when robot coordination is the primary objective. Inspired by scenario generation techniques\, we formally define the problem of Layout Optimization (LO) and propose methods to optimize layouts for maximum throughput. We then discuss our first proposed work: Multi-Objective Warehouse Layout Optimization for Throughput and Storage Capacity\, which addresses the trade-off between denser storage and efficient robot movement. We search for Pareto-optimal layouts that balance storage capacity and throughput. \nNext\, we investigate Guidance Graph Optimization in the context of multi-robot coordination in general. When robots are continuously receiving new tasks\, requiring them to move indefinitely\, existing planning algorithms inevitably make short-horizon decisions. In particular\, individually efficient paths may still create long-term congestion. To mitigate this limitation\, we introduce global movement guidance that encourages long-term cooperation among robots. We formally define the guidance graph as a representation of such guidance and the problem of Guidance Graph Optimization (GGO)\, and discuss several methods to optimize guidance graphs for throughput. However\, these methods suffer from the curse of dimensionality\, making existing GGO methods scale poorly to large graphs. Therefore\, we then discuss our second proposed work: Guidance Graph Optimization for Ultra Large Graphs\, which scales GGO methods to large graphs. \nFinally\, we address a common limitation shared by all environment optimization methods: their reliance on simplified simulation. Because realistic simulation is computationally expensive while evolutionary optimization requires many evaluations\, directly scaling these methods to realistic settings is difficult. To improve sample efficiency\, we discuss our third proposed work: Deep Surrogate Assisted Realistic Environment Optimization\, where we investigate data-driven surrogate models that approximate simulator outcomes and enable environment optimization under more realistic conditions. \nTogether\, these contributions demonstrate that optimizing physical layouts\, task assignments\, and virtual guidance can substantially improve coordination in large-scale multi-robot systems\, offering a unified perspective on environment design for scalable robotic operation. \n\n\n \n\nThesis Committee Members:\n\nJiaoyang Li (Chair)\,\nStephen Smith\,\nPeter Zhang\nStefanos Nokilaidis (University of Southern California)\n\n\nThesis Proposal Draft link: https://drive.google.com/file/d/1BwxjdPA0qOyiaOKhcuOUV4hNCcMgaTmq/view?usp=sharing
URL:https://www.ri.cmu.edu/event/ri-phd-thesis-proposal-yulun-zhang/
LOCATION:Newell-Simon Hall 3305
CATEGORIES:PhD Thesis Proposal,Student Talks
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=America/New_York:20260427T153000
DTEND;TZID=America/New_York:20260427T163000
DTSTAMP:20260707T165816
CREATED:20260421T163228Z
LAST-MODIFIED:20260421T163228Z
UID:151102-1777303800-1777307400@www.ri.cmu.edu
SUMMARY:Learning Through Fitting: Advancing Non-Pixel Representations for Visual Inference
DESCRIPTION:Abstract:  Gridded pixel and voxel representations form the backbone of visual computing\, but they struggle to scale efficiently to large\, high-dimensional data\, such as volumetric medical scans and complex scientific simulations. Consequently\, continuous\, nongridded models such as implicit neural representations (INRs) and Gaussian splatting have gained significant research traction over the past five years. However\, their use has largely been confined to signal reconstruction rather than acting as foundational data types for downstream analysis. In this talk\, I will present our recent work on elevating continuous models beyond mere signal representation. First\, I will discuss how injecting learned priors into INRs via strategic parameter initialization enables powerful new capabilities\, including rapid\, amortized fitting to novel signals and even semantic segmentation. Second\, I will briefly outline our recent efforts in performing visual recognition tasks directly on 2D Gaussian image representations. Finally\, I will highlight interesting future directions in this “learning through fitting” paradigm of visual computing. \nBio:  Guha Balakrishnan is an Assistant Professor in the Electrical and Computer Engineering Department at Rice University. His research group tackles a diverse range of problems across computer vision and imaging\, with a primary focus on developing efficient neural representations for complex visual signals and advancing responsible AI through uncertainty estimation and interpretability techniques. He frequently grounds these methods in real-world applications by collaborating with domain experts in scientific disciplines such as medicine and the geosciences. His scientific contributions have been recognized with several honors\, including the NSF CAREER Award and the MICCAI Best Paper Award. Before joining Rice\, he completed his Ph.D. at MIT’s Computer Science and Artificial Intelligence Laboratory (CSAIL)\, and earned his undergraduate degrees in Computer Science and Computer Engineering from the University of Michigan\, Ann Arbor. \nHomepage:  www.guhabalakrishnan.com \nSponsor:\nThe VASC seminar is generously sponsored by HeyGen\, an all-in-one AI-powered video generation platform that leverages advances in computer vision\, generative modeling\, and multimodal learning to make high-quality video creation both scalable and accessible.
URL:https://www.ri.cmu.edu/event/learning-through-fitting-advancing-non-pixel-representations-for-visual-inference/
LOCATION:Newell-Simon Hall 4305
CATEGORIES:Seminar,VASC Seminar
ATTACH;FMTTYPE=image/jpeg:https://www.ri.cmu.edu/app/uploads/2026/04/4-27-26-Balakrishnan.jpeg
END:VEVENT
BEGIN:VEVENT
DTSTART;TZID=America/New_York:20260427T143000
DTEND;TZID=America/New_York:20260427T153000
DTSTAMP:20260707T165816
CREATED:20260423T172923Z
LAST-MODIFIED:20260423T172923Z
UID:151117-1777300200-1777303800@www.ri.cmu.edu
SUMMARY:Solid Knitting
DESCRIPTION:Abstract:\nIn this talk\, I introduce solid knitting\, a fabrication technique that combines the layer-by-layer volumetric approach of 3D printing with the topologically entwined stitch structure of knitting to produce solid 3D objects. This technique has the potential to be 3D printing that can be unraveled. I define the basic building stitch blocks of solid knitting and demonstrate a working prototype of a solid knitting machine controlled by a low-level instruction language\, along with a volumetric design tool for creating knittable patterns. \nWhile conventional knitting has a row-column structure\, solid knitting has a row-column-layer structure. Our machine autonomously produces solid-knitted prisms\, although it requires manual intervention in the first and final layers. Our design tool allows users to create solid knitting patterns by connecting stitch blocks; objects designed in the tool can be exported as a sequence of instructions for fabrication on the machine. \nI also discuss the evaluation of the solid knitting machine\, the mechanical errors that I have encountered\, potential extensions to the capability of our machine\, and the usability of the design tool. \nCommittee:\nGabrielle Ohlson\nJames McCann\nMelisa Orta Martinez\nScott Hudson
URL:https://www.ri.cmu.edu/event/solid-knitting/
LOCATION:GHC 4303
CATEGORIES:PhD Speaking Qualifier,Student Talks
END:VEVENT
END:VCALENDAR