Planning Task Shielding: Detecting and Repairing Flaws in Planning Tasks through Turning them Unsolvable

Alberto Pozanco; Pietro Totis; Marianela Morales; Daniel Borrajo

cs.AIApr 8, 2026

Planning Task Shielding: Detecting and Repairing Flaws in Planning Tasks through Turning them Unsolvable

Authors: Alberto Pozanco, Pietro Totis, Marianela Morales, Daniel Borrajo

Organizations: J.P. Morgan AI Research

Abstract

Most research in planning focuses on generating a plan to achieve a desired set of goals. However, a goal specification can also be used to encode a property that should never hold, allowing a planner to identify a trace that would reach a flawed state. In such cases, the objective may shift to modifying the planning task to ensure that the flawed state is never reached, in other words, to make the planning task unsolvable. In this paper we introduce planning task shielding: the problem of detecting and repairing flaws in planning tasks. We propose three different approaches to solve these tasks, each aiming to minimize the number of modifications to the original actions that render the planning task unsolvable.

Figures & tables

Listing 1 Approval workflow described as PDDL actions.

Figure 1: aga succeeds on all instances with faster runtime,

\textsc{allmin}^{100}

and

\textsc{dfs}^{15}

follow.

Figure 2: Experimental results on the 90 benchmark problems. Each column denotes one of the 10 problems for each domain.

Figure 3: Execution time split into the time to compute the set of plans with symk (blue) and the time to compute the minimum number of modifications with the MILP (orange).

Figure 4: Number of modifications per benchmark and method. While aga successfully renders unsolvable all benchmarks, the average number of modifications is higher.

Figure 5: Mean execution time per benchmark and method. aga is consistently faster, while mcts shows the higher variance across the search-based methods.

Config	SYNTH4					SYNTH8					SYNTH16
	T	C	S	$\|M\|$	Time	T	C	S	$\|M\|$	Time	T	C	S	$\|M\|$	Time
allmin 10	$10$	$10$	$\mathbf{10}$	$\mathbf{6.0\pm 0.0}$	$1.1\pm 0.3$	$10$	$10$	$0$	-	-	$10$	$10$	$0$	-	-
allmin 100	$10$	$10$	$\mathbf{10}$	$\mathbf{6.0\pm 0.0}$	$1.1\pm 0.1$	$10$	$10$	$\mathbf{10}$	$\mathbf{11.0\pm 0.0}$	$5.6\pm 0.7$	$10$	$10$	$\mathbf{10}$	$\mathbf{21.0\pm 0.0}$	$69.6\pm 10.1$
allmin	$10$	$10$	$\mathbf{10}$	$\mathbf{6.0\pm 0.0}$	$1.1\pm 0.1$	$10$	$10$	$\mathbf{10}$	$\mathbf{11.0\pm 0.0}$	$5.6\pm 0.8$	$10$	$10$	$\mathbf{10}$	$\mathbf{21.0\pm 0.0}$	$68.6\pm 8.7$
aga	$10$	$10$	$\mathbf{10}$	$8.0\pm 0.0$	$\mathbf{0.6\pm 0.2}$	$10$	$10$	$\mathbf{10}$	$16.0\pm 0.0$	$\mathbf{1.1\pm 0.2}$	$10$	$10$	$\mathbf{10}$	$32.0\pm 0.0$	$\mathbf{2.7\pm 0.5}$
bfs	$10$	$10$	$\mathbf{10}$	$\mathbf{6.0\pm 0.0}$	$134.2\pm 23.7$	$10$	$0$	$0$	-	-	$10$	$0$	$0$	-	-

Table 1: Outcome breakdown and modification counts per method and benchmark. T= total problems; C=completed, S=success. T-C=out of memory or timeout, C-S=modifications do not render the task unsolvable.

Petri Net Relaxation for Infeasibility Explanation and Sequential Task Planning
Feb 25, 2026Nguyen Cong Nhat Le, John G. Rogers, Claire N. Bonial +1Sequential Decision Making Reachability Analysis
Colorado School of Mines, Golden CO 80401, USA · DEVCOM Army Research Laboratory, Adelphi, MD, USA
Counterfactual Reasoning in Automated Planning
May 4, 2026Alberto Pozanco, Daniel Borrajo, Manuela VelosoCounterfactual Reasoning
J.P. Morgan AI Research
LeanPlan: Optimal Planning with LLM-Generated Heuristics and Admissibility Proofs
Oct 6, 2026André G. Pereira, Augusto B. Corrêa, Felipe Meneguzzi +1Automated Heuristic Design LLM Planning
Federal University of Rio Grande do Sul Brazil · University of Oxford United Kingdom · University of Aberdeen United Kingdom +1

Feb 25, 2026J/K move · Enter open · S save

Petri Net Relaxation for Infeasibility Explanation and Sequential Task Planning

Nguyen Cong Nhat Le, John G. Rogers, Claire N. Bonial, Neil T. Dantam

Colorado School of Mines, Golden CO 80401, USA · DEVCOM Army Research Laboratory, Adelphi, MD, USA

Sequential Decision Making Reachability Analysis

Open paper AskPDF arXiv

Plans often change due to changes in the situation or our understanding of the situation. Sometimes, a feasible plan may not even exist, and identifying such infeasibilities is useful to determine when requirements need adjustment. Common planning approaches focus on efficient one-shot planning in feasible cases rather than updating domains or detecting infeasibility. We propose a Petri net reachability relaxation to enable robust invariant synthesis, efficient goal-unreachability detection, and helpful infeasibility explanations. We further leverage incremental constraint solvers to support goal and constraint updates. Empirically, compared to baselines, our system produces a comparable number of invariants, detects up to 2 times more infeasibilities, performs competitively in one-shot planning, and outperforms in sequential plan updates in the tested domains.

Abstract

Explore similar work

Petri Net Relaxation for Infeasibility Explanation and Sequential Task Planning

Counterfactual Reasoning in Automated Planning

LeanPlan: Optimal Planning with LLM-Generated Heuristics and Admissibility Proofs