{"id":{"repo_id":"uiuc","oai_identifier":"oai:www.ideals.illinois.edu:2142/129472"},"canonical_url":"https://search.dev.ndltd.org/etd/uiuc/oai:www.ideals.illinois.edu:2142/129472","repository":{"repo_id":"uiuc","name":"University of Illinois - Urbana-Champaign","base_url":"https://www.ideals.illinois.edu/oai-pmh"},"display":{"title":"Real-world autonomy in uncertain and unknown environments","abstract":"Submission original under an indefinite embargo labeled 'Open Access'. The submission was exported from vireo on 2025-10-19 without embargo terms","abstract_html":"Submission original under an indefinite embargo labeled 'Open Access'. The submission was exported from vireo on 2025-10-19 without embargo terms","abstract_has_math":false,"creators":["Thangeda, Pranay"],"institution":"University of Illinois Urbana-Champaign","degree_name":"Ph.D.","degree_level":"Dissertation","degree_discipline":"Aerospace Engineering","degree_department":null,"school":null,"contributors":["Ornik, Melkior","Tran, Huy T.","Allison, James","Hauser, Kris K."],"advisors":[],"committee_chairs":[],"committee_members":[],"year":2025,"date_issued":"2025-05-01","date_published":"2025-05-01","updated_at":"2026-07-22T22:25:05Z","subjects":["Autonomous Systems","Robotics","Space Robotics","Mars Rover Navigation","Planetary Rovers","Unmanned Systems","Artificial Intelligence (AI)","Machine Learning","Reinforcement Learning","Efficient Learning","Accelerated Learning","Planning under Uncertainty","Decision-Theoretic Planning","Markov Decision Processes (MDPs)","Stochastic Control","Unknown Environments","Stochastic Environments","Unknown Dynamics","Side Information Integration","Indirect Sampling","Sample Efficiency","Exploration-Exploitation","Bayesian Exploration","Value of Information","Agent Safety","Safety Guarantees","Safe Reinforcement Learning","Optimal Routing","Stochastic Travel Times","Spatial Correlations","Multiobjective Optimization","Onboard Sensors","Real-time Data"],"languages":["en","eng"],"rights":["Copyright 2025 Pranay Thangeda"],"rights_urls":[],"identifier_entries":[]},"links":{"outbound_url":"https://hdl.handle.net/2142/129472","outbound_label":"Handle","outbound_source":"dc:identifier"},"metadata_groups":[{"id":"people","label":"People","entries":[{"key":"dc:contributor","label":"Contributor","values":["Ornik, Melkior","Tran, Huy T.","Allison, James","Hauser, Kris K."]},{"key":"dc:creator","label":"Author","values":["Thangeda, Pranay"]}]},{"id":"academic_context","label":"Academic Context","entries":[{"key":"dc:date","label":"Dc Date","values":["2025-05-01","2025-05"]},{"key":"dc:type","label":"Dc Type","values":["text","Thesis"]},{"key":"thesis:degree_discipline","label":"Discipline","values":["Aerospace Engineering"]},{"key":"thesis:degree_level","label":"Degree Level","values":["Dissertation"]},{"key":"thesis:degree_name","label":"Degree Name","values":["Ph.D."]},{"key":"thesis:institution_name","label":"Thesis Institution Name","values":["University of Illinois Urbana-Champaign"]}]},{"id":"subjects_keywords","label":"Subjects and Keywords","entries":[{"key":"dc:subject","label":"Dc Subject","values":["Autonomous Systems","Robotics","Space Robotics","Mars Rover Navigation","Planetary Rovers","Unmanned Systems","Artificial Intelligence (AI)","Machine Learning","Reinforcement Learning","Efficient Learning","Accelerated Learning","Planning under Uncertainty","Decision-Theoretic Planning","Markov Decision Processes (MDPs)","Stochastic Control","Unknown Environments","Stochastic Environments","Unknown Dynamics","Side Information Integration","Indirect Sampling","Sample Efficiency","Exploration-Exploitation","Bayesian Exploration","Value of Information","Agent Safety","Safety Guarantees","Safe Reinforcement Learning","Optimal Routing","Stochastic Travel Times","Spatial Correlations","Multiobjective Optimization","Onboard Sensors","Real-time Data"]}]},{"id":"language_rights","label":"Language and Rights","entries":[{"key":"dc:language","label":"Dc Language","values":["en","eng"]},{"key":"dc:rights","label":"Dc Rights","values":["Copyright 2025 Pranay Thangeda"]}]},{"id":"identifiers","label":"Identifiers","entries":[{"key":"dc:identifier","label":"Identifier","values":["https://hdl.handle.net/2142/129472"]}]},{"id":"additional","label":"Additional Metadata","entries":[{"key":"dc:description","label":"Description","values":["Submission original under an indefinite embargo labeled 'Open Access'. The submission was exported from vireo on 2025-10-19 without embargo terms","The student, Pranay Thangeda, accepted the attached license on 2025-05-01 at 02:21.","The student, Pranay Thangeda, submitted this Dissertation for approval on 2025-05-01 at 02:32.","This Dissertation was approved for publication on 2025-05-01 at 14:41.","DSpace SAF Submission Ingestion Package generated from Vireo submission #22141 on 2025-10-19 at 18:19:44","Real-world autonomy in uncertain and unknown environments presents fundamental challenges that go beyond the controlled settings where most robotic systems are developed and tested. This dissertation develops a comprehensive framework for achieving robust autonomy by strategically leveraging the structure of different domains---from completely unknown environments requiring rapid online adaptation to partially known systems amenable to proactive planning under uncertainty. The first part of my research tackles the challenge of learning for deployment-time adaptation in completely unknown environments. I introduce a novel meta-learning approach that enables robots to adapt to drastically different granular materials with minimal online data by explicitly training for large domain shifts. Building on this foundation, I develop a closed-loop control system that fuses visual motion field with force feedback to enable dynamic adaptation during individual scooping actions. For extraterrestrial exploration, I present an adaptive sampling strategy that balances information gathering with operational constraints and risk of failure. Finally, I address continual learning in unknown environments through an adaptation strategy that enables both rapid adaptation to new domains and retention of knowledge about previously encountered scenarios. The second part explores learning and planning for uncertain deployment-time configurations, where some prior knowledge exists but specific conditions remain unknown until deployment. For crop management, I develop a constrained reinforcement learning approach that explicitly incorporates operational restrictions while handling environmental uncertainty. I scale these concepts to large agricultural order fulfillment through a hybrid tree search algorithm that combines domain knowledge with online planning. Finally, I introduce InfraLib, a comprehensive framework for modeling and managing large-scale infrastructure systems, demonstrating how appropriate computational tools can enable learning-based approaches for practical deployment challenges at massive scale. My research demonstrates that effective real-world autonomy requires carefully balancing between simulation, problem structure exploitation, and online adaptation based on the specific characteristics of the domain. The methods are validated through extensive experiments in physical testbeds and real-world case studies across domains ranging from robotic manipulation to infrastructure management. This dissertation advances our understanding of how to develop autonomous systems that can operate reliably in real-world settings by appropriately leveraging available knowledge while maintaining the ability to adapt to unknown conditions."]},{"key":"dc:format","label":"Dc Format","values":["application/pdf"]},{"key":"dc:title","label":"Title","values":["Real-world autonomy in uncertain and unknown environments"]}]}],"canonical_facts":{"dc:contributor":["Ornik, Melkior","Tran, Huy T.","Allison, James","Hauser, Kris K."],"dc:creator":["Thangeda, Pranay"],"dc:date":["2025-05-01","2025-05"],"dc:description":["Submission original under an indefinite embargo labeled 'Open Access'. The submission was exported from vireo on 2025-10-19 without embargo terms","The student, Pranay Thangeda, accepted the attached license on 2025-05-01 at 02:21.","The student, Pranay Thangeda, submitted this Dissertation for approval on 2025-05-01 at 02:32.","This Dissertation was approved for publication on 2025-05-01 at 14:41.","DSpace SAF Submission Ingestion Package generated from Vireo submission #22141 on 2025-10-19 at 18:19:44","Real-world autonomy in uncertain and unknown environments presents fundamental challenges that go beyond the controlled settings where most robotic systems are developed and tested. This dissertation develops a comprehensive framework for achieving robust autonomy by strategically leveraging the structure of different domains---from completely unknown environments requiring rapid online adaptation to partially known systems amenable to proactive planning under uncertainty. The first part of my research tackles the challenge of learning for deployment-time adaptation in completely unknown environments. I introduce a novel meta-learning approach that enables robots to adapt to drastically different granular materials with minimal online data by explicitly training for large domain shifts. Building on this foundation, I develop a closed-loop control system that fuses visual motion field with force feedback to enable dynamic adaptation during individual scooping actions. For extraterrestrial exploration, I present an adaptive sampling strategy that balances information gathering with operational constraints and risk of failure. Finally, I address continual learning in unknown environments through an adaptation strategy that enables both rapid adaptation to new domains and retention of knowledge about previously encountered scenarios. The second part explores learning and planning for uncertain deployment-time configurations, where some prior knowledge exists but specific conditions remain unknown until deployment. For crop management, I develop a constrained reinforcement learning approach that explicitly incorporates operational restrictions while handling environmental uncertainty. I scale these concepts to large agricultural order fulfillment through a hybrid tree search algorithm that combines domain knowledge with online planning. Finally, I introduce InfraLib, a comprehensive framework for modeling and managing large-scale infrastructure systems, demonstrating how appropriate computational tools can enable learning-based approaches for practical deployment challenges at massive scale. My research demonstrates that effective real-world autonomy requires carefully balancing between simulation, problem structure exploitation, and online adaptation based on the specific characteristics of the domain. The methods are validated through extensive experiments in physical testbeds and real-world case studies across domains ranging from robotic manipulation to infrastructure management. This dissertation advances our understanding of how to develop autonomous systems that can operate reliably in real-world settings by appropriately leveraging available knowledge while maintaining the ability to adapt to unknown conditions."],"dc:format":["application/pdf"],"dc:identifier":["https://hdl.handle.net/2142/129472"],"dc:language":["en","eng"],"dc:rights":["Copyright 2025 Pranay Thangeda"],"dc:subject":["Autonomous Systems","Robotics","Space Robotics","Mars Rover Navigation","Planetary Rovers","Unmanned Systems","Artificial Intelligence (AI)","Machine Learning","Reinforcement Learning","Efficient Learning","Accelerated Learning","Planning under Uncertainty","Decision-Theoretic Planning","Markov Decision Processes (MDPs)","Stochastic Control","Unknown Environments","Stochastic Environments","Unknown Dynamics","Side Information Integration","Indirect Sampling","Sample Efficiency","Exploration-Exploitation","Bayesian Exploration","Value of Information","Agent Safety","Safety Guarantees","Safe Reinforcement Learning","Optimal Routing","Stochastic Travel Times","Spatial Correlations","Multiobjective Optimization","Onboard Sensors","Real-time Data"],"dc:title":["Real-world autonomy in uncertain and unknown environments"],"dc:type":["text","Thesis"],"thesis:degree_discipline":["Aerospace Engineering"],"thesis:degree_level":["Dissertation"],"thesis:degree_name":["Ph.D."],"thesis:institution_name":["University of Illinois Urbana-Champaign"]},"updated_at":"2026-07-22T22:25:05Z"}