Roderick Bloem
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Research Subjects
Roderick Bloem has a total of 160 co-authors. Here are the top co-authors by number of shared publications:
Barbara Jobstmann
École Polytechnique Fédérale de Lausanne
13 shared publicationsRobert Könighofer
Graz University of Technology, Graz, Austria
13 shared publicationsGeorg Hofferek
Institute for Applied Information Processing and Communications (IAIK), Graz University of Technology, Austria
12 shared publications- Bettina Könighofer
Graz University of Technology
10 shared publications - Swen Jacobs
CISPA Helmholtz Center for Information Security
10 shared publications Fabio Somenzi
University of Colorado
8 shared publications
Latest Publications
- Open Access
- 8 January 2026
- Published by Association for Computing Machinery (ACM) in Proceedings of the ACM on Programming Languages
- Vol. 10 (POPL) , 2439-2467
Parameterized Infinite-State Reactive Synthesis
We propose a method to synthesize a parameterized infinite-state system that can be instantiated for different parameter values. The specification is given in a parameterized temporal logic that allows for data variables as well as parameters that encode properties of the environment. Our synthesis method runs in a counterexample-guided loop consisting of four steps: (1) we synthesize concrete systems for some small parameter instantiations using existing techniques. (2) We generalize the concrete systems into a parameterized program. (3) We create a proof candidate consisting of an invariant and a ranking function. (4) We check the proof candidate for consistency with the program. If the proof succeeds, the parameterized program is valid. Otherwise, we identify a parameter value for which it fails and add a new concrete instance to step one. To generalize programs and create proof candidates, we use a combination of anti-unification and syntax-guided synthesis to express the differences between the programs as a function of the parameters. We evaluate our approach on new examples and examples from the literature that are manually parameterized.
- 8 January 2026
- Published by Association for Computing Machinery (ACM) in Proceedings of the ACM on Programming Languages
- Vol. 10 (POPL) , 2439-2467
Open AccessParameterized Infinite-State Reactive Synthesis
We propose a method to synthesize a parameterized infinite-state system that can be instantiated for different parameter values. The specification is given in a parameterized temporal logic that allows for data variables as well as parameters that encode properties of the environment. Our synthesis method runs in a counterexample-guided loop consisting of four steps: (1) we synthesize concrete systems for some small parameter instantiations using existing techniques. (2) We generalize the concrete systems into a parameterized program. (3) We create a proof candidate consisting of an invariant and a ranking function. (4) We check the proof candidate for consistency with the program. If the proof succeeds, the parameterized program is valid. Otherwise, we identify a parameter value for which it fails and add a new concrete instance to step one. To generalize programs and create proof candidates, we use a combination of anti-unification and syntax-guided synthesis to express the differences between the programs as a function of the parameters. We evaluate our approach on new examples and examples from the literature that are manually parameterized.
- Open Access
- 29 October 2025
- Published by Association for Computing Machinery (ACM) in Communications of the ACM
- Vol. 68 (11) , 80-90
Shields for Safe Reinforcement Learning
Reinforcement learning (RL) is a prominent machine learning technique used to optimize an agent’s performance in potentially unknown environments. Despite its popularity and success, RL lacks safety guarantees, both during the learning phase and deployment. This paper reviews a runtime enforcement method called shielding that ensures provable safety for RL. We describe the underlying models, the types of guarantees that can be delivered, and the process of computing shields. Furthermore, we describe several techniques for integrating shields into RL, discuss the advantages and potential drawbacks of this integration, and highlight the current challenges in shielded learning.
- 29 October 2025
- Published by Association for Computing Machinery (ACM) in Communications of the ACM
- Vol. 68 (11) , 80-90
Open AccessShields for Safe Reinforcement Learning
Reinforcement learning (RL) is a prominent machine learning technique used to optimize an agent’s performance in potentially unknown environments. Despite its popularity and success, RL lacks safety guarantees, both during the learning phase and deployment. This paper reviews a runtime enforcement method called shielding that ensures provable safety for RL. We describe the underlying models, the types of guarantees that can be delivered, and the process of computing shields. Furthermore, we describe several techniques for integrating shields into RL, discuss the advantages and potential drawbacks of this integration, and highlight the current challenges in shielded learning.
- Open Access
Parameterized Infinite-State Reactive Synthesis
We propose a method to synthesize a parameterized infinite-state systems that can be instantiated for different parameter values. The specification is given in a parameterized temporal logic that allows for data variables as well as parameter variables that encode properties of the environment. Our synthesis method runs in a counterexample-guided loop consisting of four main steps: First, we use existing techniques to synthesize concrete systems for some small parameter instantiations. Second, we generalize the concrete systems into a parameterized program. Third, we create a proof candidate consisting of an invariant and a ranking function. Fourth, we check the proof candidate for consistency with the program. If the proof succeeds, the parameterized program is valid. Otherwise, we identify a parameter value for which the proof fails and add a new concrete instance to step one. To generalize programs and create proof candidates, we use a combination of anti-unification and syntax-guided synthesis to express syntactic differences between programs as functions of the parameters. We evaluate our approach on examples from the literature that have been extended with parameters as well as new problems.
Open AccessParameterized Infinite-State Reactive Synthesis
We propose a method to synthesize a parameterized infinite-state systems that can be instantiated for different parameter values. The specification is given in a parameterized temporal logic that allows for data variables as well as parameter variables that encode properties of the environment. Our synthesis method runs in a counterexample-guided loop consisting of four main steps: First, we use existing techniques to synthesize concrete systems for some small parameter instantiations. Second, we generalize the concrete systems into a parameterized program. Third, we create a proof candidate consisting of an invariant and a ranking function. Fourth, we check the proof candidate for consistency with the program. If the proof succeeds, the parameterized program is valid. Otherwise, we identify a parameter value for which the proof fails and add a new concrete instance to step one. To generalize programs and create proof candidates, we use a combination of anti-unification and syntax-guided synthesis to express syntactic differences between programs as functions of the parameters. We evaluate our approach on examples from the literature that have been extended with parameters as well as new problems.
- Open Access
- 4 February 2025
- Published by Springer Nature in International Journal on Software Tools for Technology Transfer
- p. 1-19
An adaptive, provable correct simplex architecture
Simplex architectures optimize performance and safety by switching between an advanced controller and a base controller. We propose an approach to synthesize the switching logic and extensions of the base controller in the Simplex architectures to achieve high performance and provable correctness for a rich class of temporal specifications by maximizing the time the advanced controller is active. We achieve provable correctness by performing static verification of the baseline controller. The result of this verification is a set of states that is proven to be safe, called the recoverable region. We employ proofs on demand to ensure that the base controller is safe in those states that are visited during runtime, which depends on the advanced controller. Verification of hybrid systems is often overly conservative, resulting in smaller recoverable regions that cause unnecessary switches to the baseline controller. To avoid these switches, we invoke targeted reachability queries to extend the recoverable region at runtime. In case the recoverable region cannot be extended using the baseline controller, we employ a repair procedure. This tries to synthesize a patch for the baseline controller and can further extend the recoverable region. Our offline and online verification relies upon reachability analysis since it allows observation-based extension of the known recoverable region. We implemented our methodology on top of the state-of-the-art tool HyPro which allowed us to automatically synthesize verified and performant Simplex architectures for advanced case studies, like safe autonomous driving on a race track.
- 4 February 2025
- Published by Springer Nature in International Journal on Software Tools for Technology Transfer
- p. 1-19
Open AccessAn adaptive, provable correct simplex architecture
Simplex architectures optimize performance and safety by switching between an advanced controller and a base controller. We propose an approach to synthesize the switching logic and extensions of the base controller in the Simplex architectures to achieve high performance and provable correctness for a rich class of temporal specifications by maximizing the time the advanced controller is active. We achieve provable correctness by performing static verification of the baseline controller. The result of this verification is a set of states that is proven to be safe, called the recoverable region. We employ proofs on demand to ensure that the base controller is safe in those states that are visited during runtime, which depends on the advanced controller. Verification of hybrid systems is often overly conservative, resulting in smaller recoverable regions that cause unnecessary switches to the baseline controller. To avoid these switches, we invoke targeted reachability queries to extend the recoverable region at runtime. In case the recoverable region cannot be extended using the baseline controller, we employ a repair procedure. This tries to synthesize a patch for the baseline controller and can further extend the recoverable region. Our offline and online verification relies upon reachability analysis since it allows observation-based extension of the known recoverable region. We implemented our methodology on top of the state-of-the-art tool HyPro which allowed us to automatically synthesize verified and performant Simplex architectures for advanced case studies, like safe autonomous driving on a race track.
- 23 January 2025
- Published by Springer Nature
- p. 137-159
Synthesis of Controllers for Continuous Blackbox Systems
- 23 January 2025
- Published by Springer Nature
- p. 137-159
Synthesis of Controllers for Continuous Blackbox Systems
- 26 October 2024
- Published by Springer Nature
- p. 281-301
Synthesis from Infinite-State Generalized Reactivity(1) Specifications
- 26 October 2024
- Published by Springer Nature
- p. 281-301
Synthesis from Infinite-State Generalized Reactivity(1) Specifications
- Open Access
Safety Shielding under Delayed Observation
Agents operating in physical environments need to be able to handle delays in the input and output signals since neither data transmission nor sensing or actuating the environment are instantaneous. Shields are correct-by-construction runtime enforcers that guarantee safe execution by correcting any action that may cause a violation of a formal safety specification. Besides providing safety guarantees, shields should interfere minimally with the agent. Therefore, shields should pick the safe corrective actions in such a way that future interferences are most likely minimized. Current shielding approaches do not consider possible delays in the input signals in their safety analyses. In this paper, we address this issue. We propose synthesis algorithms to compute \emph{delay-resilient shields} that guarantee safety under worst-case assumptions on the delays of the input signals. We also introduce novel heuristics for deciding between multiple corrective actions, designed to minimize future shield interferences caused by delays. As a further contribution, we present the first integration of shields in a realistic driving simulator. We implemented our delayed shields in the driving simulator \textsc{Carla}. We shield potentially unsafe autonomous driving agents in different safety-critical scenarios and show the effect of delays on the safety analysis.
Open AccessSafety Shielding under Delayed Observation
Agents operating in physical environments need to be able to handle delays in the input and output signals since neither data transmission nor sensing or actuating the environment are instantaneous. Shields are correct-by-construction runtime enforcers that guarantee safe execution by correcting any action that may cause a violation of a formal safety specification. Besides providing safety guarantees, shields should interfere minimally with the agent. Therefore, shields should pick the safe corrective actions in such a way that future interferences are most likely minimized. Current shielding approaches do not consider possible delays in the input signals in their safety analyses. In this paper, we address this issue. We propose synthesis algorithms to compute \emph{delay-resilient shields} that guarantee safety under worst-case assumptions on the delays of the input signals. We also introduce novel heuristics for deciding between multiple corrective actions, designed to minimize future shield interferences caused by delays. As a further contribution, we present the first integration of shields in a realistic driving simulator. We implemented our delayed shields in the driving simulator \textsc{Carla}. We shield potentially unsafe autonomous driving agents in different safety-critical scenarios and show the effect of delays on the safety analysis.
- Open Access
- 1 July 2023
- Published by Association for the Advancement of Artificial Intelligence (AAAI) in Proceedings of the International Conference on Automated Planning and Scheduling
- Vol. 33 (1) , 80-85
Safety Shielding under Delayed Observation
Agents operating in physical environments need to be able to handle delays in the input and output signals since neither data transmission nor sensing or actuating the environment are instantaneous. Shields are correct-by-construction runtime enforcers that guarantee safe execution by correcting any action that may cause a violation of a formal safety specification. Besides providing safety guarantees, shields should interfere minimally with the agent. Therefore, shields should pick the safe corrective actions in such a way that future interferences are most likely minimized. Current shielding approaches do not consider possible delays in the input signals in their safety analyses. In this paper, we address this issue. We propose synthesis algorithms to compute delay-resilient shields that guarantee safety under worst-case assumptions on the delays of the input signals. We also introduce novel heuristics for deciding between multiple corrective actions, designed to minimize future shield interferences caused by delays. As a further contribution, we present the first integration of shields in a realistic driving simulator. We implemented our delayed shields in the driving simulator Carla. We shield potentially unsafe autonomous driving agents in different safety-critical scenarios and show the effect of delays on the safety analysis.
- 1 July 2023
- Published by Association for the Advancement of Artificial Intelligence (AAAI) in Proceedings of the International Conference on Automated Planning and Scheduling
- Vol. 33 (1) , 80-85
Open AccessSafety Shielding under Delayed Observation
Agents operating in physical environments need to be able to handle delays in the input and output signals since neither data transmission nor sensing or actuating the environment are instantaneous. Shields are correct-by-construction runtime enforcers that guarantee safe execution by correcting any action that may cause a violation of a formal safety specification. Besides providing safety guarantees, shields should interfere minimally with the agent. Therefore, shields should pick the safe corrective actions in such a way that future interferences are most likely minimized. Current shielding approaches do not consider possible delays in the input signals in their safety analyses. In this paper, we address this issue. We propose synthesis algorithms to compute delay-resilient shields that guarantee safety under worst-case assumptions on the delays of the input signals. We also introduce novel heuristics for deciding between multiple corrective actions, designed to minimize future shield interferences caused by delays. As a further contribution, we present the first integration of shields in a realistic driving simulator. We implemented our delayed shields in the driving simulator Carla. We shield potentially unsafe autonomous driving agents in different safety-critical scenarios and show the effect of delays on the safety analysis.
- 2 May 2023
- Published by Springer Nature
- p. 141-160
Provable Correct and Adaptive Simplex Architecture for Bounded-Liveness Properties
- 2 May 2023
- Published by Springer Nature
- p. 141-160
Provable Correct and Adaptive Simplex Architecture for Bounded-Liveness Properties
- 29 December 2022
- Published by Springer Nature
- p. 650-663
Correct-by-Construction Runtime Enforcement in AI – A Survey
- 29 December 2022
- Published by Springer Nature
- p. 650-663
Correct-by-Construction Runtime Enforcement in AI – A Survey