{"id":{"repo_id":"usf","oai_identifier":"oai:digitalcommons.usf.edu:etd-10940"},"canonical_url":"https://search.ndltd.org/etd/usf/oai:digitalcommons.usf.edu:etd-10940","repository":{"repo_id":"usf","name":"University of South Florida","base_url":"https://digitalcommons.usf.edu/do/oai/"},"display":{"title":"Stochastic Models for Resilience Assessment and Improvement","abstract":"Resilience refers to the ability of a system to absorb and mitigate the impact of potential disruptions and return to normal operational conditions. The above notion of resilience allows us to distinguish two stages in a system's post-disruption response, the absorptive dimension and the recovery dimension. The absorptive dimension is related to a system's robustness and capacity to mitigate initial loss posterior to a disruption. Meanwhile, the recovery dimension is the system's rapidity to return and reach an acceptable level of functionality after the disruption occurrence. A social-physical (SP) system's post-disruption response is related to its resilience capacity so building more resilient systems is the key to managing and dealing with unexpected disruptions. However, many resilience actions typically involve economic impacts, which limit their implementation. This problem has been documented in different contexts. For example, during the early stages of pandemic mitigation, strict lockdowns are effective against virus propagation but may lead to harmful economic effects in the communities. Therefore, finding a balance between being resilient and the costs of being resilient is essential. This dissertation proposes implementing stochastic dynamic optimization models based on an infinite horizon Continuous-Time Markov Decision Processes to balance the intervention investment and enhance an SP system's overall resilience. We do so by reducing the initial drop in performance and the recovery time during the post-disruption response. Since the system's evolution is progressive and involves uncertainty, we can model the system response as a Markov process. In the models' formulation, we use common elements in resilience literature to provide a broader framework that facilitates the application in different disruption scenarios. For our state-space definition, we use the idea that in high-magnitude events with long recovery times, the recovery process can be modeled by considering multiple performance intervals. On the other hand, actions and rewards are related to the investment required to guarantee a transition between states. We assess the model's performance using testbeds based on different disruption scenarios, such as natural disasters and pandemic outbreaks. The proposed model can provide policymakers with information on the optimal recovery strategy and how to allocate resilience resources during the post-disruption response.","abstract_html":"Resilience refers to the ability of a system to absorb and mitigate the impact of potential disruptions and return to normal operational conditions. The above notion of resilience allows us to distinguish two stages in a system&#x27;s post-disruption response, the absorptive dimension and the recovery dimension. The absorptive dimension is related to a system&#x27;s robustness and capacity to mitigate initial loss posterior to a disruption. Meanwhile, the recovery dimension is the system&#x27;s rapidity to return and reach an acceptable level of functionality after the disruption occurrence. A social-physical (SP) system&#x27;s post-disruption response is related to its resilience capacity so building more resilient systems is the key to managing and dealing with unexpected disruptions. However, many resilience actions typically involve economic impacts, which limit their implementation. This problem has been documented in different contexts. For example, during the early stages of pandemic mitigation, strict lockdowns are effective against virus propagation but may lead to harmful economic effects in the communities. Therefore, finding a balance between being resilient and the costs of being resilient is essential. This dissertation proposes implementing stochastic dynamic optimization models based on an infinite horizon Continuous-Time Markov Decision Processes to balance the intervention investment and enhance an SP system&#x27;s overall resilience. We do so by reducing the initial drop in performance and the recovery time during the post-disruption response. Since the system&#x27;s evolution is progressive and involves uncertainty, we can model the system response as a Markov process. In the models&#x27; formulation, we use common elements in resilience literature to provide a broader framework that facilitates the application in different disruption scenarios. For our state-space definition, we use the idea that in high-magnitude events with long recovery times, the recovery process can be modeled by considering multiple performance intervals. On the other hand, actions and rewards are related to the investment required to guarantee a transition between states. We assess the model&#x27;s performance using testbeds based on different disruption scenarios, such as natural disasters and pandemic outbreaks. The proposed model can provide policymakers with information on the optimal recovery strategy and how to allocate resilience resources during the post-disruption response.","abstract_has_math":false,"creators":["Ardila Rueda, Weimar"],"institution":"University of South Florida","degree_name":"Doctor of Philosophy (Ph.D.)","degree_level":"Dissertation","degree_discipline":"Industrial and Management Systems Engineering","degree_department":null,"school":null,"contributors":["Alex Savachkin, Ph.D.","Walter Silva-Sotillo, Ph.D.","Kingsley Reeves, Ph.D.","Alex Volinsky, Ph.D.","Daniel Romero-Rodriguez, Ph.D."],"advisors":[],"committee_chairs":[],"committee_members":[],"year":2022,"date_issued":"2022-11-21T08:00:00Z","date_published":"2022-11-21T08:00:00Z","updated_at":"2026-11-19T08:48:36Z","subjects":["Dynamic Programming","Markov Decision Process","Post-Disruption Response","Resilience Enhacement","Resource Allocation","Industrial Engineering"],"languages":[],"rights":[],"rights_urls":[],"identifier_entries":[],"person_identifier_entries":[]},"links":{"outbound_url":"https://digitalcommons.usf.edu/etd/9743","outbound_label":"Repository record","outbound_source":"dc:identifier","record_url":"https://digitalcommons.usf.edu/etd/9743","record_label":"Repository record","record_kind":"record","record_source":"dc:identifier","download_url":null,"download_label":null,"download_source":null,"access_status":null,"embargo_until":null,"access_note":null},"source_record":{"url":"https://digitalcommons.usf.edu/do/oai/?verb=GetRecord&metadataPrefix=oai_etdms&identifier=oai%3Adigitalcommons.usf.edu%3Aetd-10940","prefix":"oai_etdms"},"metadata_groups":[{"id":"people","label":"People","entries":[{"key":"dc:contributor","label":"Contributor","values":["Alex Savachkin, Ph.D.","Walter Silva-Sotillo, Ph.D.","Kingsley Reeves, Ph.D.","Alex Volinsky, Ph.D.","Daniel Romero-Rodriguez, Ph.D."]},{"key":"dc:creator","label":"Author","values":["Ardila Rueda, Weimar"]}]},{"id":"academic_context","label":"Academic Context","entries":[{"key":"dc:date","label":"Dc Date","values":["2022-11-21T08:00:00Z"]},{"key":"thesis:degree_discipline","label":"Discipline","values":["Industrial and Management Systems Engineering"]},{"key":"thesis:degree_level","label":"Degree Level","values":["Dissertation","Ph.D."]},{"key":"thesis:degree_name","label":"Degree Name","values":["Doctor of Philosophy (Ph.D.)"]},{"key":"thesis:institution_name","label":"Thesis Institution Name","values":["University of South Florida"]}]},{"id":"subjects_keywords","label":"Subjects and Keywords","entries":[{"key":"dc:subject","label":"Dc Subject","values":["Dynamic Programming","Markov Decision Process","Post-Disruption Response","Resilience Enhacement","Resource Allocation","Industrial Engineering"]}]},{"id":"identifiers","label":"Identifiers","entries":[{"key":"dc:identifier","label":"Identifier","values":["https://digitalcommons.usf.edu/etd/9743"]}]},{"id":"additional","label":"Additional Metadata","entries":[{"key":"dc:description","label":"Description","values":["Resilience refers to the ability of a system to absorb and mitigate the impact of potential disruptions and return to normal operational conditions. The above notion of resilience allows us to distinguish two stages in a system's post-disruption response, the absorptive dimension and the recovery dimension. The absorptive dimension is related to a system's robustness and capacity to mitigate initial loss posterior to a disruption. Meanwhile, the recovery dimension is the system's rapidity to return and reach an acceptable level of functionality after the disruption occurrence. A social-physical (SP) system's post-disruption response is related to its resilience capacity so building more resilient systems is the key to managing and dealing with unexpected disruptions. However, many resilience actions typically involve economic impacts, which limit their implementation. This problem has been documented in different contexts. For example, during the early stages of pandemic mitigation, strict lockdowns are effective against virus propagation but may lead to harmful economic effects in the communities. Therefore, finding a balance between being resilient and the costs of being resilient is essential. This dissertation proposes implementing stochastic dynamic optimization models based on an infinite horizon Continuous-Time Markov Decision Processes to balance the intervention investment and enhance an SP system's overall resilience. We do so by reducing the initial drop in performance and the recovery time during the post-disruption response. Since the system's evolution is progressive and involves uncertainty, we can model the system response as a Markov process. In the models' formulation, we use common elements in resilience literature to provide a broader framework that facilitates the application in different disruption scenarios. For our state-space definition, we use the idea that in high-magnitude events with long recovery times, the recovery process can be modeled by considering multiple performance intervals. On the other hand, actions and rewards are related to the investment required to guarantee a transition between states. We assess the model's performance using testbeds based on different disruption scenarios, such as natural disasters and pandemic outbreaks. The proposed model can provide policymakers with information on the optimal recovery strategy and how to allocate resilience resources during the post-disruption response."]},{"key":"dc:title","label":"Title","values":["Stochastic Models for Resilience Assessment and Improvement"]}]}],"canonical_facts":{"dc:contributor":["Alex Savachkin, Ph.D.","Walter Silva-Sotillo, Ph.D.","Kingsley Reeves, Ph.D.","Alex Volinsky, Ph.D.","Daniel Romero-Rodriguez, Ph.D."],"dc:creator":["Ardila Rueda, Weimar"],"dc:date":["2022-11-21T08:00:00Z"],"dc:description":["Resilience refers to the ability of a system to absorb and mitigate the impact of potential disruptions and return to normal operational conditions. The above notion of resilience allows us to distinguish two stages in a system's post-disruption response, the absorptive dimension and the recovery dimension. The absorptive dimension is related to a system's robustness and capacity to mitigate initial loss posterior to a disruption. Meanwhile, the recovery dimension is the system's rapidity to return and reach an acceptable level of functionality after the disruption occurrence. A social-physical (SP) system's post-disruption response is related to its resilience capacity so building more resilient systems is the key to managing and dealing with unexpected disruptions. However, many resilience actions typically involve economic impacts, which limit their implementation. This problem has been documented in different contexts. For example, during the early stages of pandemic mitigation, strict lockdowns are effective against virus propagation but may lead to harmful economic effects in the communities. Therefore, finding a balance between being resilient and the costs of being resilient is essential. This dissertation proposes implementing stochastic dynamic optimization models based on an infinite horizon Continuous-Time Markov Decision Processes to balance the intervention investment and enhance an SP system's overall resilience. We do so by reducing the initial drop in performance and the recovery time during the post-disruption response. Since the system's evolution is progressive and involves uncertainty, we can model the system response as a Markov process. In the models' formulation, we use common elements in resilience literature to provide a broader framework that facilitates the application in different disruption scenarios. For our state-space definition, we use the idea that in high-magnitude events with long recovery times, the recovery process can be modeled by considering multiple performance intervals. On the other hand, actions and rewards are related to the investment required to guarantee a transition between states. We assess the model's performance using testbeds based on different disruption scenarios, such as natural disasters and pandemic outbreaks. The proposed model can provide policymakers with information on the optimal recovery strategy and how to allocate resilience resources during the post-disruption response."],"dc:identifier":["https://digitalcommons.usf.edu/etd/9743"],"dc:subject":["Dynamic Programming","Markov Decision Process","Post-Disruption Response","Resilience Enhacement","Resource Allocation","Industrial Engineering"],"dc:title":["Stochastic Models for Resilience Assessment and Improvement"],"thesis:degree_discipline":["Industrial and Management Systems Engineering"],"thesis:degree_level":["Dissertation","Ph.D."],"thesis:degree_name":["Doctor of Philosophy (Ph.D.)"],"thesis:institution_name":["University of South Florida"]},"updated_at":"2026-11-19T08:48:36Z"}