Synthesis for Robots: Guarantees and Feedback for Robot Behavior

Kress-Gazit, Hadas; Lahijanian, Morteza; Raman, Vasumathi

doi:10.1146/annurev-control-060117-104838

Cited by 119 publications

(64 citation statements)

References 113 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Furthermore, specific approaches to the synthesis of controllers, that are realized in a compositional manner, have been proposed in [ZA18] and [ZPMT11]. A survey on controller synthesis for robotics can be found in [KGLR18].…”

Section: Multi-player Support For Modeling and Reasoning On Robotic Amentioning

confidence: 99%

PuRSUE -from specification of robotic environments to synthesis of controllers

et al. 2020

View full text Add to dashboard Cite

Developing robotic applications is a complex task, which requires skills that are usually only possessed by highly-qualified robotic developers. While formal methods that help developers in the creation and design of robotic applications exist, they must be explicitly customized to be impactful in the robotics domain and to support effectively the growth of the robotic market. Specifically, the robotic market is asking for techniques that: (i) enable a systematic and rigorous design of robotic applications though high-level languages; and (ii) enable the automatic synthesis of low-level controllers, which allow robots to achieve their missions. To address these problems we present the PuRSUE (Planner for RobotS in Uncontrollable Environments) approach, which aims to support developers in the rigorous and systematic design of high-level run-time control strategies for robotic applications. The approach includes PuRSUE-ML a high-level language that allows for modeling the environment, the agents deployed therein, and their missions. PuRSUE is able to check automatically whether a controller that allows robots to achieve their missions might exist and, then, it synthesizes a controller. We evaluated how PuRSUE helps designers in modeling robotic applications, the effectiveness of its automatic computation of controllers, and how the approach supports the deployment of controllers on actual robots. The evaluation is based on 13 scenarios derived from 3 different robotic applications presented in the literature. The results show that: (i) PuRSUE-ML is effective in supporting designers in the formal modeling of robotic applications compared to a direct encoding of robotic applications in low-level modeling formalisms; (ii) PuRSUE enables the automatic generation of controllers that are difficult to create manually; and (iii) the plans generated with PuRSUE are indeed effective when deployed on actual robots.

show abstract

Section: Multi-player Support For Modeling and Reasoning On Robotic Amentioning

confidence: 99%

PuRSUE -from specification of robotic environments to synthesis of controllers

et al. 2020

View full text Add to dashboard Cite

show abstract

“…Control and program synthesis are techniques to automatically transform high-level specifications into control or programs guaranteed to satisfy the specification. In robotics, researchers typically use different temporal logics to express tasks and automatically transform them into robot behaviors [55]. Thus, users can reason about the robot's overall task rather than implementation details.…”

Section: Control Synthesismentioning

confidence: 99%

“…LTL allows users to encode assumptions about the behavior of the robot's environment (e.g., the state of the PwMCI) and requirements on the robot behavior (e.g., if the PwMCI is not engaged, play music). Furthermore, there exist algorithms that automatically transform an LTL formula into a finite state controller [55] that is then used for robot control. For computational reasons, we use the GR(1) fragment of LTL [14] as the underlying formalism.…”

Section: Control Synthesismentioning

confidence: 99%

Jessie

Kubota

Peterson

Rajendren

et al. 2020

Proceedings of the 2020 ACM/IEEE International Conference on Human-Robot Interaction

Self Cite

View full text Add to dashboard Cite

JESSIE is a robotic system that enables novice programmers to program social robots by expressing high-level specifications. We employ control synthesis with a tangible front-end to allow users to define complex behavior for which we automatically generate control code. We demonstrate JESSIE in the context of enabling clinicians to create personalized treatments for people with mild cognitive impairment (MCI) on a Kuri robot, in little time and without error. We evaluated JESSIE with neuropsychologists who reported high usability and learnability. They gave suggestions for improvement, including increased support for personalization, multi-party programming, collaborative goal setting, and re-tasking robot role post-deployment, which each raise technical and sociotechnical issues in HRI. We exhibit JESSIE's reproducibility by replicating a clinician-created program on a TurtleBot 2. As an open-source means of accessing control synthesis, JESSIE supports reproducibility, scalability, and accessibility of personalized robots for HRI. CCS CONCEPTS • Computer systems organization → Robotic control; • Applied computing → Health informatics.

show abstract

“…There have been various variants of LTL developed over the years for different reasons. We describe a few here, for a more detailed treatment please refer KressGazit et al [25]. To handle uncertainty, a probabilistic version of LTL, Probabilistic Computation Tree Logic [16] has been defined so as to evaluate over states of an MDP.…”

Section: Related Workmentioning

confidence: 99%

“…Although we choose GLTL paired with an MDP to find policies corresponding to LTL expressions, our language grounding system can work with any framework for computing a satisfying controller for the robot, such as those described in the related work [25]. Notice that the switch is commensurate with defining a new machine translation problem with a target language defined by the syntax of the alternative framework.…”

Section: B Geometric Linear Temporal Logic (Gltl)mentioning

confidence: 99%

Sequence-to-Sequence Language Grounding of Non-Markovian Task Specifications

Gopalan

Arumugam

Wong

et al. 2018

Robotics: Science and Systems XIV

View full text Add to dashboard Cite

Abstract-Often times, natural language commands issued to robots not only specify a particular target configuration or goal state but also outline constraints on how the robot goes about its execution. That is, the path taken to achieving some goal state is given equal importance to the goal state itself. One example of this could be instructing a wheeled robot to "go to the living room but avoid the kitchen," in order to avoid scuffing the floor. This class of behaviors poses a serious obstacle to existing language understanding for robotics approaches that map to either action sequences or goal state representations. Due to the non-Markovian nature of the objective, approaches in the former category must map to potentially unbounded action sequences whereas approaches in the latter category would require folding the entirety of a robot's trajectory into a (traditionally Markovian) state representation, resulting in an intractable decision-making problem. To resolve this challenge, we use a recently introduced probabilistic variant of Linear Temporal Logic (LTL) as a goal specification language for a Markov Decision Process (MDP). While demonstrating that standard neural sequence-to-sequence learning models can successfully ground language to this semantic representation, we also provide analysis that highlights generalization to novel, unseen logical forms as an open problem for this class of model. We evaluate our system within two simulated robot domains as well as on a physical robot, demonstrating accurate language grounding alongside a significant expansion in the space of interpretable robot behaviors.

show abstract

Synthesis for Robots: Guarantees and Feedback for Robot Behavior

Cited by 119 publications

References 113 publications

PuRSUE -from specification of robotic environments to synthesis of controllers

PuRSUE -from specification of robotic environments to synthesis of controllers

Jessie

Sequence-to-Sequence Language Grounding of Non-Markovian Task Specifications

Contact Info

Product

Resources

About