arXiv subjects

Bettina Könighofer

Publications and source records attributed to Bettina Könighofer.

23 records · Page 2Linked to original sources

arXiv2019

Safe Reinforcement Learning via Probabilistic Shields

This paper targets the efficient construction of a safety shield for decision making in scenarios that incorporate uncertainty. Markov decision processes (MDPs) are prominent models to capture such planning problems. Reinforcement learning (RL) is a machine learning technique to determine near-optimal policies in MDPs that may be unknown prior to exploring the model. However, during exploration, RL is prone to induce behavior that is undesirable or not allowed in safety- or mission-critical contexts. We introduce the concept of a probabilistic shield that enables decision-making to adhere to safety constraints with high probability. In a separation of concerns, we employ formal verification to efficiently compute the probabilities of critical decisions within a safety-relevant fragment of the MDP. We use these results to realize a shield that is applied to an RL algorithm which then optimizes the actual performance objective. We discuss tradeoffs between sufficient progress in exploration of the environment and ensuring safety. In our experiments, we demonstrate on the arcade game PAC-MAN and on a case study involving service robots that the learning efficiency increases as the learning needs orders of magnitude fewer episodes.

cs.AI↗

arXiv2019

Synthesis of Admissible Shields

Shield synthesis is an approach to enforce a set of safety-critical properties of a reactive system at runtime. A shield monitors the system and corrects any erroneous output values instantaneously. The shield deviates from the given outputs as little as it can and recovers to hand back control to the system as soon as possible. This paper takes its inspiration from a case study on mission planning for unmanned aerial vehicles (UAVs) in which k-stabilizing shields, which guarantee recovery in a finite time, could not be constructed. We introduce the notion of admissible shields, which improves k-stabilizing shields in two ways: (1) whereas k-stabilizing shields take an adversarial view on the system, admissible shields take a collaborative view. That is, if there is no shield that guarantees recovery within k steps regardless of system behavior, the admissible shield will attempt to work with the system to recover as soon as possible. (2) Admissible shields can handle system failures during the recovery phase. In our experimental results we show that for UAVs, we can generate admissible shields, even when k-stabilizing shields do not exist.

cs.LO↗

arXiv2017

Safe Reinforcement Learning via Shielding

Reinforcement learning algorithms discover policies that maximize reward, but do not necessarily guarantee safety during learning or execution phases. We introduce a new approach to learn optimal policies while enforcing properties expressed in temporal logic. To this end, given the temporal logic specification that is to be obeyed by the learning system, we propose to synthesize a reactive system called a shield. The shield is introduced in the traditional learning process in two alternative ways, depending on the location at which the shield is implemented. In the first one, the shield acts each time the learning agent is about to make a decision and provides a list of safe actions. In the second way, the shield is introduced after the learning agent. The shield monitors the actions from the learner and corrects them only if the chosen action causes a violation of the specification. We discuss which requirements a shield must meet to preserve the convergence guarantees of the learner. Finally, we demonstrate the versatility of our approach on several challenging reinforcement learning scenarios.

cs.LO↗

arXiv2013

Synthesizing Multiple Boolean Functions using Interpolation on a Single Proof

It is often difficult to correctly implement a Boolean controller for a complex system, especially when concurrency is involved. Yet, it may be easy to formally specify a controller. For instance, for a pipelined processor it suffices to state that the visible behavior of the pipelined system should be identical to a non-pipelined reference system (Burch-Dill paradigm). We present a novel procedure to efficiently synthesize multiple Boolean control signals from a specification given as a quantified first-order formula (with a specific quantifier structure). Our approach uses uninterpreted functions to abstract details of the design. We construct an unsatisfiable SMT formula from the given specification. Then, from just one proof of unsatisfiability, we use a variant of Craig interpolation to compute multiple coordinated interpolants that implement the Boolean control signals. Our method avoids iterative learning and back-substitution of the control functions. We applied our approach to synthesize a controller for a simple two-stage pipelined processor, and present first experimental results.

cs.LO↗

arXiv2012

Synthesizing Robust Systems with RATSY

Specifications for reactive systems often consist of environment assumptions and system guarantees. An implementation should not only be correct, but also robust in the sense that it behaves reasonably even when the assumptions are (temporarily) violated. We present an extension of the requirements analysis and synthesis tool RATSY that is able to synthesize robust systems from GR(1) specifications, i.e., system in which a finite number of safety assumption violations is guaranteed to induce only a finite number of safety guarantee violations. We show how the specification can be turned into a two-pair Streett game, and how a winning strategy corresponding to a correct and robust implementation can be computed. Finally, we provide some experimental results.

cs.LO↗