Pith. sign in

REVIEW 4 major objections 5 minor 39 references

An Iterative Approach for Heterogeneous Multi-Agent Route Planning with Resource Transportation Uncertainty and Temporal Logic Goals

T0 review · 4 major / 5 minor · reviewed 2026-08-05 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗

Pith's one-line read This paper claims that a heterogeneous robot team can satisfy a Capability Temporal Logic mission with resource transportation even when the resource map is initially unknown, by alternating between a MILP planner that maximizes partial sat

desk verdict A plausible iterative explore-exploit planner for CaTL under unknown resource maps, but the '100% satisfaction' headline is planning-objective convergence on a belief map, not executed mission success. read the letter →

arxiv 2508.19429 v1 pith:HQVV4JDI submitted 2025-08-26 cs.RO cs.FL

Gustavo A. Cardona , Kaier Liang , Cristian-Ioan Vasile This is my paper
classification cs.ROcs.FL
keywords multi-agentrouteplanningCapabilityTemporalLogic(CaTL)resourceuncertaintyexploration-exploitationbalanceKalmanfilterbeliefupdatemixed-integerlinearprogrammingheterogeneousrobotteams
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

This paper tries to show that a heterogeneous robot team can carry out a mission specified in Capability Temporal Logic even when it starts with zero knowledge of where resources are located. The key idea is an iterative loop: each round the team solves a mixed-integer linear program that maximizes partial mission satisfaction plus an exploration bonus for visiting uncertain locations, executes the plan, then updates a Kalman-filter belief about resource amounts. In the reported simulation, the satisfaction fraction rises from 0 to 100 percent in seven iterations as the resource-estimation error falls. If the approach holds up, it gives a practical way to plan transport and capability tasks under uncertainty without knowing the resource map in advance.

What carries the argument

The loop couples two maps: the belief resource map R̃, initialized to zero, stores the estimated amount of each resource at each location, and a Kalman filter updates it from observations; the covariance matrix yields a normalized uncertainty map Ω. The MILP maximizes J_satisfaction + αJ_exploration, where J_satisfaction is a recursive partial-satisfaction encoding of the CaTL formula translated to STL, and J_exploration = Σ y(q)Ω(q) rewards visiting high-uncertainty locations. The exploration weight α decays each iteration, shifting the planner from information gathering to task exploitation as the belief converges.

What would settle it

Deliberately mis-specify the sensor noise in the Kalman filter (for example, set the assumed standard deviation half or double the true value) and measure whether the satisfaction fraction still converges to 100 percent; or run back-to-back missions where the resource distribution changes between iterations rather than resetting, and check whether convergence stalls or the exploration term targets irrelevant locations.

Watch

Extended reading notes

Core claim

This paper claims that a team of heterogeneous robots can satisfy a CaTL mission with transportation and capability constraints even when the initial resource map is completely unknown, by iterating between planning and estimation. At each iteration the planner solves a MILP whose objective is the partial-satisfaction fraction of the specification plus a weighted exploration term that rewards visits to locations with high estimated variance; after execution, a Kalman filter fuses sensor readings into a belief map and an uncertainty map. The authors report that this loop drives the satisfaction fraction from 0 to 100 percent in seven iterations in the main case study, and that across their sc

Load-bearing premise

The loop assumes the team is given an observation model whose noise statistics match the real sensors, so the Kalman-filter belief updates are unbiased; it also assumes the resource distribution resets after each mission, so a learned belief stays valid for the next execution.

Editorial extensions

If this is right

  • A mission can start with an empty resource survey: the first plan satisfies as much of the specification as the current belief allows, and later plans refine as the belief map improves.
  • The planner automatically balances exploration against task progress, so the team does not need a separate exploration phase or a hand-designed information-gathering heuristic.
  • In the reported experiments, convergence to 100 percent satisfaction happens in a small number of iterations (around 7 in the main case study) across different grid sizes and agent counts.
  • Because the satisfaction objective is a partial-satisfaction encoding, the optimizer produces a numeric satisfaction fraction between 0 and 1, which is what lets the loop monitor convergence.
  • The approach inherits the MILP encoding of CaTL with resource constraints, so capability, quantity, and deadline requirements remain encoded in the plan up to the accuracy of the current belief.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • The exploration reward is variance-based and mission-agnostic, so a mission-aware reward that weights uncertainty only where resources are needed could cut wasted visits; that is an extension the paper does not test.
  • The reported convergence depends on resources resetting each mission; if resources are consumed or move between iterations, the Kalman-filter update would need a process model or forgetting factor.
  • The same alternating structure transfers to other unknowns, such as travel durations or sensor noise, by swapping in the corresponding estimator—a pattern consistent with the authors' earlier travel-uncertainty work.
  • The 100-percent results are simulated with noise parameters used by construction; a field test with real sensors would show whether the assumed observation model is accurate enough to reproduce convergence.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

4 major / 5 minor

Summary. The paper considers heterogeneous multi-robot route planning under Capability Temporal Logic (CaTL) specifications when the resource distribution in the environment is initially unknown. The authors propose an iterative algorithm that alternates between (i) solving a MILP that maximizes a weighted sum of partial mission satisfaction and an exploration objective based on an uncertainty map, and (ii) updating a Gaussian belief over resource amounts with a Kalman filter from observations collected during plan execution. The central claim is that this iterative scheme makes the satisfaction fraction converge to 100% in a small number of iterations, and that the approach scales with environment size and agent count. The evaluation consists of one convergence experiment and a scalability study over grid sizes and agent counts, with run-time and iteration counts reported as averages over five runs.

Significance. If the claims were fully supported, the paper would address a practically relevant gap: coordinating heterogeneous teams under formal temporal-logic specifications when resource availability is uncertain and only discovered over time. The idea of combining partial-satisfaction MILP planning with a Kalman-filter belief refinement is sensible and the motivating planetary-exploration scenario is well chosen. The paper also builds on a credible prior line of work on CaTL and partial satisfaction encodings. However, the evidence presented is thin and partly ambiguous: the headline convergence result is based on a single run, no baselines are compared, the MILP encoding is not stated in sufficient detail for reproduction, and the reported 'satisfaction fraction' appears to be a planning-level quantity rather than the satisfaction actually achieved during execution. Because the latter issue affects the interpretation of the paper's main claim, the result is not yet established at the level expected for a journal publication, although the deficiencies are addressable within the scope of the manuscript.

major comments (4)
  1. [Section IV, Fig. 3(a); Section III (after Alg. 1); Section II-A] The central claim that 'the satisfaction fraction steadily converges to 100 percent in 7 iterations' is supported by the planning objective J_satisfaction computed from the current belief map, not by the satisfaction realized during execution. The paper states in Section II-A that agents pick up resources 'according to the belief of that state, except when the true value is less than the belief,' and in Section III explicitly says the approach does 'not consider mission failure or online adaptation due to the overestimation of required resources during the task execution.' Thus, if the Kalman-filter belief overestimates a resource at a pickup location, the executed plan will acquire fewer resources than planned and the CaTL task may fail. Fig. 3(a) alone is therefore not evidence that executed missions achieve 100% satisfaction; it may simply reflect convergence of the belief to the true
  2. [Section IV, Figs. 3 and 5] The evaluation is too thin to support the convergence and scalability claims as stated. The main convergence result (Fig. 3) is a single run with no statistical spread, no random seeds, and no variation in the underlying resource distribution. The scalability study reports averages of five runs without error bars or variance information, and no baselines are compared (e.g., no iterative planning without the exploration term, no random exploration, no non-iterative one-shot planning). The termination condition 'when the satisfaction fraction converges' is also not formally defined. Without multiple trials, a formal convergence criterion, and at least one baseline, the reader cannot assess whether the observed behavior is due to the algorithm or to the specific problem instance and by-construction convergence of the belief map.
  3. [Section III-A, Eq. (7)] The MILP that is central to the method is not fully specified. Equation (7) states only the objective 'max J_satisfaction + alpha J_exploration' and the constraints are referenced to prior work ('robots and resources dynamics [23]', 'partial satisfaction encoding of phi [32]'). While the exploration constraints (5)-(6) are given, the full integer programming formulation—including resource transportation dynamics, capacity constraints, pickup decisions under the belief map, and the recursive partial-satisfaction encoding—is omitted. This makes the empirical results irreproducible and makes it difficult to judge whether the execution rule in Section II-A is actually enforced in the planned trajectories. The authors should include the complete encoding, at least in an appendix or supplementary material, or provide enough detail to reconstruct it without access to the cited prior papers.
  4. [Section III-B, Eq. (8)] The belief-update component relies on the assumption that 'an observation model is provided such that a Kalman Filter can be applied.' The observation model and the Kalman update equations are not defined, and the sensor standard deviations in the simulations are set by construction. Because the uncertainty map Ω in Eq. (8) directly determines the exploration objective (4), a misspecified observation model could bias exploration and, together with the overestimation issue above, degrade executed satisfaction. The paper should provide a robustness analysis, or at least a sensitivity study with intentionally mismatched noise parameters, to substantiate the claim that the method is robust under sensor uncertainty. At a minimum, the observation model should be stated explicitly.
minor comments (5)
  1. [Section IV, scalability experiment] The task description has a likely typo: 'requires all type a and half of type c agents move to top right corner and all type b and half of type c agents to the right bottom corner' appears swapped relative to the definitions of phi1 (Bottom Right) and phi2 (Top Right). Also, the expression '(c,⌊M, 2⌋)' in phi2 is garbled; it should probably be ⌊M/2⌋ or ⌈M/2⌉.
  2. [Section III-B] The notation for the camera capability is confusing: c0 is described both as a capability and as a vector of standard deviations σ. Clarify the relationship between a robot's sensor capability and the measurement noise parameters.
  3. [Section IV] The termination criterion 'when the satisfaction fraction converges' should be made precise (e.g., tolerance and number of consecutive iterations). Without a definition, the iteration counts in Fig. 5(b) are not reproducible.
  4. [Remark 1] The assumption that resources reset at the start of each task execution means the learned belief map remains valid across iterations. This assumption is stated but not discussed in terms of practical limitations; it would be helpful to comment on how resource depletion or production would break the current convergence interpretation.
  5. [Throughout] Several minor typographical issues: '◊' is used interchangeably with '◇' for the eventually operator; the simulation specification uses a fullwidth colon in the time intervals; and the references to the authors' own prior work, particularly [23] and [32], are frequent but not always sufficiently contextualized for a reader who wants to isolate the novel contribution of this paper.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the central claim is an empirical demonstration, and the self-citations used are component encodings, not premises that contain the conclusion.

full rationale

The paper proposes an iterative planner and reports simulation results; it does not claim a first-principles derivation of its convergence. The plotted 'satisfaction fraction' (Sec. IV, Fig. 3a) is the MILP objective J_satisfaction from Eq. (7), i.e., the degree to which the plan satisfies the CaTL specification under the current belief map R~. Reporting that this optimized quantity reaches 100% once the Kalman-filter belief converges to the true resource map is an empirical outcome of the optimizer, not a quantity that is equal to its inputs by construction. The paper explicitly disclaims overestimation-induced mission failure ('We do not consider mission failure or online adaptation due to the overestimation of required resources during the task execution'), which is a validity caveat about the metric, not a circular step. The load-bearing encodings for CaTL resource transportation and partial STL satisfaction are cited to prior peer-reviewed work by the same authors ([23], [32], [33]); they are used as components with stated semantics, and the present paper's contribution—iterative exploration, belief refinement, and uncertainty-weighted exploration—does not reduce to those citations. No uniqueness theorem, ansatz, or fitted parameter is renamed as a prediction. Thus no step in the derivation chain is equivalent to its own input.

Assumptions & free parameters 2 free parameters · 4 assumptions · 0 invented entities

The method rests on the authors' earlier MILP encodings ([23], [32], [33]), a user-supplied Kalman-filter observation model, and a resetting resource distribution (Remark 1). These are all taken as inputs without independent validation in this paper.

free parameters (2)
  • exploration weight alpha = 1, decaying by factor 0.8 per iteration
    Balances mission satisfaction and exploration in the MILP objective (7); chosen by hand and decayed, with no tuning procedure or sensitivity analysis.
  • resource variance weights alpha_h = not specified
    These weights appear in the uncertainty map normalization (8) but their values are never given in the paper, leaving a quantity that directly affects the exploration objective unspecified.
assumptions (4)
  • standard math CaTL is a fragment of STL
    Used in Section III-A1 to justify translating CaTL tasks into STL formulas; this is a known result from the CaTL literature.
  • domain assumption The partial-satisfaction STL encoding from [32], [33] is correct and applicable
    The full MILP encoding is omitted and deferred to prior work by the same authors; the paper's central optimization depends on this encoding being correct.
  • domain assumption Resources reset to original levels at the start of each mission
    Remark 1 states this assumption, which ensures the belief map learned in one iteration remains relevant for the next; without it, the iterative refinement could chase a moving target.
  • domain assumption A Kalman-filter observation model is provided and accurate
    Section III-B assumes an observation model is given; the belief update and uncertainty map (8) depend entirely on this model and its noise parameters.

how reviews work

0 comments
Cite this review

Pith. "Pith review of An Iterative Approach for Heterogeneous Multi-Agent Route Planning with Resource Transportation Uncertainty and Temporal Logic Goals." pith.science (2026). https://pith.science/paper/HQVV4JDI

@misc{pith2026250819429,
  author       = {Pith},
  title        = {Pith review of: An Iterative Approach for Heterogeneous Multi-Agent Route Planning with Resource Transportation Uncertainty and Temporal Logic Goals},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/HQVV4JDI}},
  note         = {Machine review of arXiv:2508.19429}
}
read the original abstract

This paper presents an iterative approach for heterogeneous multi-agent route planning in environments with unknown resource distributions. We focus on a team of robots with diverse capabilities tasked with executing missions specified using Capability Temporal Logic (CaTL), a formal framework built on Signal Temporal Logic to handle spatial, temporal, capability, and resource constraints. The key challenge arises from the uncertainty in the initial distribution and quantity of resources in the environment. To address this, we introduce an iterative algorithm that dynamically balances exploration and task fulfillment. Robots are guided to explore the environment, identifying resource locations and quantities while progressively refining their understanding of the resource landscape. At the same time, they aim to maximally satisfy the mission objectives based on the current information, adapting their strategies as new data is uncovered. This approach provides a robust solution for planning in dynamic, resource-constrained environments, enabling efficient coordination of heterogeneous teams even under conditions of uncertainty. Our method's effectiveness and performance are demonstrated through simulated case studies.

Figures

Figures reproduced from arXiv: 2508.19429 by the authors.

Figure 1
Figure 1. Example of heterogeneous robot planning on a [PITH_FULL_IMAGE:figures/full_fig_p001_1.png] view at source ↗
Figure 2
Figure 2. Simulation Environment and c ∶ (bottom right, 3). Two types of resources, h1 and h2, are distributed in certain nodes, each resource with a total amount of 50 in the environment, ensuring that maximum satisfaction to 100 percent is feasible. The specification is given as: ϕ = ◊[0∶30]ϕ1 ∧ ◊[0∶30]ϕ2 ∧ ◊[0∶30]ϕ3 ϕ1 = (2,red, (a, 1), (b, 1), (h1, 15), (h2, 15)) ϕ2 = (1,yellow, (c, 2), (a, 1), (b, 1), (h1, 15), (h2, 15))… view at source ↗
Figure 3
Figure 3. Iterative performance (a) satisfaction fraction, (b) [PITH_FULL_IMAGE:figures/full_fig_p006_3.png] view at source ↗
Figures from the paper (1 more)
Figure 5
Figure 5. Figure 5: Scalability analysis (a) run time performance, [PITH_FULL_IMAGE:figures/full_fig_p006_5.png]

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

39 extracted references · 38 canonical work pages

  1. [23]

    Planning for heterogeneous teams of robots with temporal logic, capability, and resource constraints,

    G. A. Cardona and C.-I. Vasile, “Planning for heterogeneous teams of robots with temporal logic, capability, and resource constraints,” The International Journal of Robotics Research , p. 02783649241247285

    work page

  2. [32]

    Partial Satisfaction of Signal Tempo- ral Logic Specifications for Coordination of Multi-robot Systems,

    G. A. Cardona and C.-I. Vasile, “Partial Satisfaction of Signal Tempo- ral Logic Specifications for Coordination of Multi-robot Systems,” in Workshop on the Algorithmic Foundations of Robotics , pp. 223–238, Springer, 2022

    work page 2022

  3. [1]

    Robot swarm navigation and victim detection using rendezvous consensus in search and rescue operations,

    G. A. Cardona and J. M. Calderon, “Robot swarm navigation and victim detection using rendezvous consensus in search and rescue operations,” Applied Sciences, vol. 9, no. 8, p. 1702, 2019

    work page 2019

  4. [2]

    Application of multi-robot systems to disaster-relief scenarios with limited communication,

    J. Gregory, J. Fink, E. Stump, J. Twigg, J. Rogers, D. Baran, N. Fung, and S. Young, “Application of multi-robot systems to disaster-relief scenarios with limited communication,” in Field and Service Robotics: Results of the 10th International Conference , pp. 639–653, Springer, 2016

    work page 2016

  5. [3]

    Planning for modular aerial robotic tools with temporal logic constraints,

    G. A. Cardona, D. Salda ˜na, and C.-I. Vasile, “Planning for modular aerial robotic tools with temporal logic constraints,” in IEEE Confer- ence on Decision and Control (CDC) , pp. 2878–2883, IEEE, 2022

    work page 2022

  6. [4]

    Nasa robotics research for planetary surface exploration,

    C. R. Weisbin and G. Rodriguez, “Nasa robotics research for planetary surface exploration,” IEEE Robotics & Automation Magazine , vol. 7, no. 4, pp. 25–34, 2000

    work page 2000

  7. [5]

    An autonomous mobile robotic system for surveillance of indoor environments,

    D. Di Paola, A. Milella, G. Cicirelli, and A. Distante, “An autonomous mobile robotic system for surveillance of indoor environments,” Inter- national Journal of Advanced Robotic Systems , vol. 7, no. 1, p. 8, 2010

    work page 2010

  8. [6]

    Construction robotics: From automation to collabora- tion,

    S. Parascho, “Construction robotics: From automation to collabora- tion,” Annual Review of Control, Robotics, and Autonomous Systems , vol. 6, no. 1, pp. 183–204, 2023

    work page 2023

Show all 39 references
  1. [7]

    Aerial surveillance system using uav,

    Z. Zaheer, A. Usmani, E. Khan, and M. A. Qadeer, “Aerial surveillance system using uav,” in 2016 thirteenth international conference on wireless and optical communications networks (WOCN) , pp. 1–7, IEEE, 2016

    2016

  2. [8]

    Continuous execution of high-level collaborative tasks for heterogeneous robot teams,

    A. Fang, T. Yin, J. Lin, and H. Kress-Gazit, “Continuous execution of high-level collaborative tasks for heterogeneous robot teams,” arXiv preprint arXiv:2406.18019, 2024

    2024 arXiv

  3. [9]

    Accomplishing high-level tasks with modular robots,

    G. Jing, T. Tosun, M. Yim, and H. Kress-Gazit, “Accomplishing high-level tasks with modular robots,” Autonomous Robots , vol. 42, pp. 1337–1354, 2018

    2018

  4. [10]

    An end-to-end system for accomplishing tasks with modular robots.,

    G. Jing, T. Tosun, M. Yim, and H. Kress-Gazit, “An end-to-end system for accomplishing tasks with modular robots.,” in Robotics: Science and systems, vol. 2, 2016

    2016

  5. [11]

    Energy-aware planning of heterogeneous multi-agent systems for serving cooperative tasks with temporal logic specifications,

    A. T. Buyukkocak, D. Aksaray, and Y . Yazıcıo ˘glu, “Energy-aware planning of heterogeneous multi-agent systems for serving cooperative tasks with temporal logic specifications,” in 2023 IEEE/RSJ Interna- tional Conference on Intelligent Robots and Systems (IROS), pp. 8659– 86...

    2023

  6. [12]

    Multi-agent planning under local ltl specifications and event-based synchronization,

    J. Tumova and D. V . Dimarogonas, “Multi-agent planning under local ltl specifications and event-based synchronization,” Automatica, vol. 70, pp. 239–248, 2016

    2016

  7. [13]

    Stylus*: A temporal logic optimal control synthesis algorithm for large-scale multi-robot systems,

    Y . Kantaros and M. M. Zavlanos, “Stylus*: A temporal logic optimal control synthesis algorithm for large-scale multi-robot systems,” The International Journal of Robotics Research , vol. 39, no. 7, pp. 812– 836, 2020

    2020

  8. [14]

    Temporal-logic- based reactive mission and motion planning,

    H. Kress-Gazit, G. E. Fainekos, and G. J. Pappas, “Temporal-logic- based reactive mission and motion planning,” IEEE transactions on robotics, vol. 25, no. 6, pp. 1370–1381, 2009

    2009

  9. [15]

    Multi-agent motion plan- ning from signal temporal logic specifications,

    D. Sun, J. Chen, S. Mitra, and C. Fan, “Multi-agent motion plan- ning from signal temporal logic specifications,” IEEE Robotics and Automation Letters, vol. 7, no. 2, pp. 3451–3458, 2022

    2022

  10. [16]

    Event-based signal temporal logic synthesis for single and multi-robot tasks,

    D. Gundana and H. Kress-Gazit, “Event-based signal temporal logic synthesis for single and multi-robot tasks,” IEEE Robotics and Au- tomation Letters, vol. 6, no. 2, pp. 3687–3694, 2021

    2021

  11. [17]

    Coupled multi-robot systems under linear temporal logic and signal temporal logic tasks,

    L. Lindemann, J. Nowak, L. Sch ¨onb¨achler, M. Guo, J. Tumova, and D. V . Dimarogonas, “Coupled multi-robot systems under linear temporal logic and signal temporal logic tasks,” IEEE Transactions on Control Systems Technology , vol. 29, no. 2, pp. 858–865, 2019

    2019

  12. [18]

    Maps2: Multi- robot anytime motion planning under signal temporal logic specifica- tions,

    M. Sewlia, C. K. Verginis, and D. V . Dimarogonas, “Maps2: Multi- robot anytime motion planning under signal temporal logic specifica- tions,” arXiv preprint arXiv:2309.05632 , 2023

    2023

  13. [19]

    Planning of heterogeneous multi-agent systems under signal temporal logic spec- ifications with integral predicates,

    A. T. Buyukkocak, D. Aksaray, and Y . Yazıcıo ˘glu, “Planning of heterogeneous multi-agent systems under signal temporal logic spec- ifications with integral predicates,” IEEE Robotics and Automation Letters, vol. 6, no. 2, pp. 1375–1382, 2021

    2021

  14. [20]

    Scratchs: Scalable and robust algorithms for task- based coordination from high-level specifications,

    A. Jones, K. Leahy, C. Vasile, S. Sadraddini, Z. Serlin, R. Tron, and C. Belta, “Scratchs: Scalable and robust algorithms for task- based coordination from high-level specifications,” in International Symposium of Robotics Research , pp. 224–241, 2019

    2019

  15. [21]

    Scalable and robust algorithms for task-based coordination from high-level specifications (scratches),

    K. Leahy, Z. Serlin, C.-I. Vasile, A. Schoer, A. M. Jones, R. Tron, and C. Belta, “Scalable and robust algorithms for task-based coordination from high-level specifications (scratches),” IEEE Transactions on Robotics, vol. 38, no. 4, pp. 2516–2535, 2021

    2021

  16. [22]

    Fast decomposition of temporal logic specifications for heterogeneous teams,

    K. Leahy, A. Jones, and C.-I. Vasile, “Fast decomposition of temporal logic specifications for heterogeneous teams,” IEEE Robotics and Automation Letters, vol. 7, no. 2, pp. 2297–2304, 2022

    2022

  17. [24]

    Safe control under uncertainty with prob- abilistic signal temporal logic,

    D. Sadigh and A. Kapoor, “Safe control under uncertainty with prob- abilistic signal temporal logic,” in Proceedings of Robotics: Science and Systems XII , 2016

    2016

  18. [25]

    Auctioning over probabilistic options for temporal logic-based multi-robot coopera- tion under uncertainty,

    P. Schillinger, M. B ¨urger, and D. V . Dimarogonas, “Auctioning over probabilistic options for temporal logic-based multi-robot coopera- tion under uncertainty,” in 2018 IEEE International Conference on Robotics and Automation (ICRA) , pp. 7330–7337, IEEE, 2018

    2018

  19. [26]

    Hierarchical motion planning under probabilistic temporal tasks and safe-return constraints,

    M. Guo, T. Liao, J. Wang, and Z. Li, “Hierarchical motion planning under probabilistic temporal tasks and safe-return constraints,” IEEE Transactions on Automatic Control , vol. 68, no. 11, pp. 6727–6742, 2023

    2023

  20. [27]

    Safe autonomy un- der perception uncertainty using chance-constrained temporal logic,

    S. Jha, V . Raman, D. Sadigh, and S. A. Seshia, “Safe autonomy un- der perception uncertainty using chance-constrained temporal logic,” Journal of Automated Reasoning , vol. 60, pp. 43–62, 2018

    2018

  21. [28]

    Probabilistic co- ordination of heterogeneous teams from capability temporal logic specifications,

    M. Cai, K. Leahy, Z. Serlin, and C.-I. Vasile, “Probabilistic co- ordination of heterogeneous teams from capability temporal logic specifications,” IEEE Robotics and Automation Letters , vol. 7, no. 2, pp. 1190–1197, 2021

    2021

  22. [29]

    An iterative approach for heterogeneous multi-agent route planning with temporal logic goals and travel duration uncertainty,

    K. Liang, G. A. Cardona, and C.-I. Vasile, “An iterative approach for heterogeneous multi-agent route planning with temporal logic goals and travel duration uncertainty,” in 2024 IEEE International Conference on Robotics and Automation (ICRA) , pp. 257–263, IEEE, 2024

    2024

  23. [30]

    Monitoring temporal properties of contin- uous signals,

    O. Maler and D. Nickovic, “Monitoring temporal properties of contin- uous signals,” in Formal Techniques, Modelling and Analysis of Timed and Fault-Tolerant Systems, pp. 152–166, Springer, 2004

    2004

  24. [31]

    Robust multi-agent coordination from catl+ specifications,

    W. Liu, K. Leahy, Z. Serlin, and C. Belta, “Robust multi-agent coordination from catl+ specifications,” in 2023 American Control Conference (ACC), pp. 3529–3534, IEEE, 2023

    2023

  25. [33]

    Preferences on partial satisfaction using weighted signal temporal logic specifications,

    G. A. Cardona and C.-I. Vasile, “Preferences on partial satisfaction using weighted signal temporal logic specifications,” in2023 European Control Conference (ECC), pp. 1–6, IEEE, 2023

    2023

  26. [34]

    Probabilistic robotics,

    S. Thrun, “Probabilistic robotics,” Communications of the ACM , vol. 45, no. 3, pp. 52–57, 2002

    2002

  27. [35]

    Gurobi optimizer reference manual,

    L. Gurobi Optimization, “Gurobi optimizer reference manual,” 2020

    2020

  28. [36]

    A flexible and efficient temporal logic tool for python: Pytelo,

    G. A. Cardona, K. Leahy, M. Mann, and C.-I. Vasile, “A flexible and efficient temporal logic tool for python: Pytelo,” arXiv preprint arXiv:2310.08714, 2023

    2023 arXiv

  29. [37]

    Parr, The definitive ANTLR reference: building domain-specific languages

    T. Parr, The definitive ANTLR reference: building domain-specific languages. Pragmatic Bookshelf, 2007

    2007

  30. [38]

    Ltl optimal multi-agent planner (lomap)

    C.-I. Vasile and A. Ulusoy, “Ltl optimal multi-agent planner (lomap).” https://github.com/wasserfeder/lomap, 2024

    2024

  31. [39]

    Exploring network structure, dynamics, and function using networkx,

    A. A. Hagberg, D. A. Schult, and P. J. Swart, “Exploring network structure, dynamics, and function using networkx,” in Proceedings of the 7th Python in Science Conference (G. Varoquaux, T. Vaught, and J. Millman, eds.), (Pasadena, CA USA), pp. 11 – 15, 2008

    2008

Pith tools

Reviewed August 5, 2026 · model on record in the stance chip above.