Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Planning and Acting While the Clock Ticks

Andrew Coles, Erez Karpas, Andrey Lavrinenko, Wheeler Ruml, Solomon Eyal Shimony, Shahaf Shperberg

TL;DR

This paper proposes a new problem setting: concurrent planning and execution, in which actions can be dispatched (executed) before planning terminates, and shows that when there is strong time pressure, the approach outperforms situated temporal planning.

Abstract

Standard temporal planning assumes that planning takes place offline and then execution starts at time 0. Recently, situated temporal planning was introduced, where planning starts at time 0 and execution occurs after planning terminates. Situated temporal planning reflects a more realistic scenario where time passes during planning. However, in situated temporal planning a complete plan must be generated before any action is executed. In some problems with time pressure, timing is too tight to complete planning before the first action must be executed. For example, an autonomous car that has a truck backing towards it should probably move out of the way now and plan how to get to its destination later. In this paper, we propose a new problem setting: concurrent planning and execution, in which actions can be dispatched (executed) before planning terminates. Unlike previous work on planning and execution, we must handle wall clock deadlines that affect action applicability and goal achievement (as in situated planning) while also supporting dispatching actions before a complete plan has been found. We extend previous work on metareasoning for situated temporal planning to develop an algorithm for this new setting. Our empirical evaluation shows that when there is strong time pressure, our approach outperforms situated temporal planning.

Planning and Acting While the Clock Ticks

TL;DR

This paper proposes a new problem setting: concurrent planning and execution, in which actions can be dispatched (executed) before planning terminates, and shows that when there is strong time pressure, the approach outperforms situated temporal planning.

Abstract

Standard temporal planning assumes that planning takes place offline and then execution starts at time 0. Recently, situated temporal planning was introduced, where planning starts at time 0 and execution occurs after planning terminates. Situated temporal planning reflects a more realistic scenario where time passes during planning. However, in situated temporal planning a complete plan must be generated before any action is executed. In some problems with time pressure, timing is too tight to complete planning before the first action must be executed. For example, an autonomous car that has a truck backing towards it should probably move out of the way now and plan how to get to its destination later. In this paper, we propose a new problem setting: concurrent planning and execution, in which actions can be dispatched (executed) before planning terminates. Unlike previous work on planning and execution, we must handle wall clock deadlines that affect action applicability and goal achievement (as in situated planning) while also supporting dispatching actions before a complete plan has been found. We extend previous work on metareasoning for situated temporal planning to develop an algorithm for this new setting. Our empirical evaluation shows that when there is strong time pressure, our approach outperforms situated temporal planning.
Paper Structure (17 sections, 3 theorems, 12 equations, 1 figure)

This paper contains 17 sections, 3 theorems, 12 equations, 1 figure.

Table of Contents

  1. Introduction
  2. Problem Statement
  3. Prior Work
  4. Abstract Metareasoning
  5. MR in Concurrent Planning & Execution
  6. Metareasoning in a Planner
  7. Metareasoning with Measurement Model
  8. The CoPEM Model
  9. Basic Tractable Case
  10. Extension to Incorporate Measurements
  11. Leveraging the Tractable Case in Practice
  12. Implementation within the Planner
  13. State-Space Modifications to Support Acting
  14. Dispatch Estimates during Search
  15. Metareasoning for Action Dispatch Decisions
  16. ...and 2 more sections

Key Result

Theorem 1

Given a CoPE problem instance $I$ and an initiation function $f$, using the effective deadlines as computation deadlines (and subsequently ignoring the $H_i$) defines a $KDS(AE)^2$ instance $KDS(I,f)$. The optimal policy for $I$ restricted to action execution according to $f$ is equal to the optima

Figures (1)

  • Figure 1: Problems solved vs. expansions-per-sec: RCLL

Theorems & Definitions (3)

  • Theorem 1
  • Theorem 2
  • Theorem 3