Loading…

Using Knowledge Awareness to improve Safety of Autonomous Driving

We present a method, which incorporates knowledge awareness into the symbolic computation of discrete controllers for reactive cyber physical systems, to improve decision making about the unknown operating environment under uncertain/incomplete inputs. Assuming an abstract model of the system and th...

Full description

Saved in:
Bibliographic Details
Published in:arXiv.org 2023-10
Main Authors: Calvagna, Andrea, Ghosh, Arabinda, Soudjani, Sadegh
Format: Article
Language:English
Subjects:
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:We present a method, which incorporates knowledge awareness into the symbolic computation of discrete controllers for reactive cyber physical systems, to improve decision making about the unknown operating environment under uncertain/incomplete inputs. Assuming an abstract model of the system and the environment, we translate the knowledge awareness of the operating context into linear temporal logic formulas and incorporate them into the system specifications to synthesize a controller. The knowledge base is built upon an ontology model of the environment objects and behavioural rules, which includes also symbolic models of partial input features. The resulting symbolic controller support smoother, early reactions, which improves the security of the system over existing approaches based on incremental symbolic perception. A motion planning case study for an autonomous vehicle has been implemented to validate the approach, and presented results show significant improvements with respect to safety of state-of-the-art symbolic controllers for reactive systems.
ISSN:2331-8422