Actions you can handle: dependent types for AI plans

@article{Hill2021ActionsYC,
  title={Actions you can handle: dependent types for AI plans},
  author={Alasdair Hill and Ekaterina Komendantskaya and Matthew L. Daggitt and Ronald P. A. Petrick},
  journal={Proceedings of the 6th ACM SIGPLAN International Workshop on Type-Driven Development},
  year={2021}
}
Verification of AI is a challenge that has engineering, algorithmic and programming language components. For example, AI planners are deployed to model actions of autonomous agents. They comprise a number of searching algorithms that, given a set of specified properties, find a sequence of actions that satisfy these properties. Although AI planners are mature tools from the algorithmic and engineering points of view, they have limitations as programming languages. Decidable and efficient… 

Figures from this paper

References

SHOWING 1-10 OF 36 REFERENCES

Proof-Carrying Plans

A formal framework and a prototype Agda implementation are presented that represent PDDL plans as executable functions that inhabit types that are given by formulae describing planning problems and automatically ensures that the generated program corresponds precisely to the specification of the planning problem.

Proof-Carrying Plans: a Resource Logic for AI Planning

A novel resource logic, the Proof Carrying Plans (PCP) logic that can be used to verify plans produced by AI planners, and a complete Agda formalisation of the PCP logic and of its soundness proof.

Classical AI Planning as Theorem Proving: The Case of a Fragment of Linear Logic

This paper attempts to evaluate the use of a theorem prover in the multiplicative fragment of linear logic which has been shown to simulate conjunctive Strips-like planning 9]. A proof search

A Formally Verified Validator for Classical Planning Problems and Solutions

This paper formalises the semantics of a fragment of PDDL (V, ¬, →, = in the preconditions, typing and constants) in the Higher-Order Logic theorem prover Isabelle/HOL and constructs an efficient plan validator and mechanically proves it correct w.r.t. the authors' semantics.

PDDL-the planning domain definition language

This manual describes the syntax of PDDL, the Planning Domain Definition Language, the problem-specification language for the AIPS-98 planning competition, and hopes to encourage empirical evaluation of planner performance, and development of standard sets of problems all in comparable notations.

Towards Explainable AI Planning as a Service

It is argued that Explainable Planning can be designed as a service -- that is, as a wrapper around an existing planning system that utilises the existing planner to assist in answering contrastive questions.

SOME PHILOSOPHICAL PROBLEMS FROM THE STANDPOINT OF ARTI CIAL INTELLIGENCE

BI as an assertion language for mutable data structures

A model in which the law of the excluded middleholds is given is given, thus showing that the approach is compatible with classical logic, and a local character enjoyed by specifications in the logic is described, which enables a class of frame axioms, which say what parts of the heap don't change, to be inferred automatically.

Local Reasoning about Programs that Alter Data Structures

An extension of Hoare's logic for reasoning about programs that alter data structures is described, based on a possible worlds model of the logic of bunched implications, and includes spatial conjunction and implication connectives alongside those of classical logic.

Logic programming in a fragment of intuitionistic linear logic

  • J. S. HodasD. Miller
  • Computer Science
    [1991] Proceedings Sixth Annual IEEE Symposium on Logic in Computer Science
  • 1991
The intuitionistic notion of context is refined by using a fragment of J.-Y. Girard's (Theor. Comput. Sci., vol.50, p.1-102, 1987) linear logic that includes additive and multiplicative conjunction,