pith. sign in

arxiv: 2606.09203 · v2 · pith:RTIQIOJBnew · submitted 2026-06-08 · 💻 cs.RO

Deterministic Execution of ROS 2 Applications via Lingua Franca

Harun Teper , Shaokai Lin , Shulu Li , Edward A. Lee , Jian-Jia Chen This is my paper

Pith reviewed 2026-06-27 16:27 UTC · model grok-4.3

classification 💻 cs.RO
keywords ROS 2deterministic executionLingua Francalogical timerobot middlewarepub-subcallback schedulingAutoware
0
0 comments X

The pith

ROS 2 applications can be automatically converted to run with deterministic callback order under Lingua Franca logical time.

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

The paper establishes a conversion process that takes an existing ROS 2 program and executes it inside Lingua Franca so the same inputs always trigger callbacks in the same order. Standard ROS 2 pub-sub leaves callback ordering nondeterministic even inside one executor, and this nondeterminism grows with network effects in distributed setups. The conversion first isolates the ROS 2 features that fit logical-time determinism, then builds an automatic translator that produces an equivalent LF program. LF mechanisms such as logical delays and federated execution can then be layered on without altering the original ROS 2 source.

Core claim

By mapping selected ROS 2 features to Lingua Franca constructs that enforce logical-time ordering, an unmodified ROS 2 application can be made to produce identical callback sequences on every run of the same input, as verified on both a synthetic case and the Autoware reference system where default ROS 2 exhibited varying orders and latencies while the LF version did not.

What carries the argument

The automatic conversion framework that extracts callback and topic structure from ROS 2 and emits an equivalent LF program while preserving observable behavior.

If this is right

  • Identical inputs produce identical execution orders.
  • End-to-end latencies remain consistent across runs.
  • Logical-time delays and fault handling from LF can be added to the original ROS 2 code.
  • Federated execution across processes becomes available while keeping the ROS 2 interface unchanged.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • Safety analysis techniques that assume deterministic ordering could now be applied directly to converted ROS 2 systems.
  • The same conversion pattern might be tested on other pub-sub robotics frameworks to check generality.
  • Distributed deployments could gain timing predictability by using LF federation on top of the converted nodes.

Load-bearing premise

Typical ROS 2 applications use only features that map cleanly to deterministic logical-time execution without loss of required semantics.

What would settle it

Execute the converted LF version of the Autoware reference system on repeated identical inputs and record differing callback execution orders across runs.

Figures

Figures reproduced from arXiv: 2606.09203 by Edward A. Lee, Harun Teper, Jian-Jia Chen, Shaokai Lin, Shulu Li.

Figure 1
Figure 1. Figure 1: Running example as a ROS 2 system. running_example A (0, 100 ms) alpha beta B alpha gamma C beta delta D 1 2 gamma delta 1 ms 1 ms 1 ms 1 ms [PITH_FULL_IMAGE:figures/full_fig_p003_1.png] view at source ↗
Figure 3
Figure 3. Figure 3: Running example: end-to-end latency distribution [PITH_FULL_IMAGE:figures/full_fig_p010_3.png] view at source ↗
Figure 4
Figure 4. Figure 4: The colcon lf pipeline: launch files, C++ source, and CMake metadata are extracted to system.json, then converted to .lf source by applying the embedding E, compiled by lfc, and executed under LF time authority. The output of the extraction tool is a system graph written to a single json file, enumerating all components that are ex￾tracted, including launch files, executables, libraries, callbacks, timers,… view at source ↗
read the original abstract

The Robot Operating System 2 (ROS 2) is a widely used middleware for robotic systems, characterized by a publish-subscribe (pub-sub) communication mechanism in which computation is structured as callbacks dispatched by ROS 2 executors. Despite its popularity, the pub-sub pattern in ROS 2 is inherently nondeterministic: the order in which these callbacks run is nondeterministic even within a single executor, and distributed deployments add further nondeterminism from the interleaving of messages across nodes and from network latency. Such nondeterminism often leads to concurrency issues and makes it virtually impossible to analyze for safeness and provide guarantees. We present a framework that is able to convert an unmodified ROS 2 application and run it under Lingua Franca (LF), a coordination language for deterministic execution using logical time, so that the same input always produces the same deterministic execution order. We first describe which ROS 2 features can be executed deterministically under logical time. Such features enable the possibility to establish an automatic conversion framework to extract information from a ROS 2 application and directly convert it into an LF program. The rich features of LF, such as logical-time delays, federated execution across processes, and fault handling, can then be applied to make the ROS 2 application be executed in a deterministic and timing-predictable manner without changing the ROS 2 code. We evaluate the framework on a synthetic example and on the Autoware reference system. We show that the order in which callbacks are executed differs in default ROS 2, while also having end-to-end latencies that vary across executions. In contrast, our LF-controlled ROS 2 system produces a deterministic execution order and consistent end-to-end latencies.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

0 major / 2 minor

Summary. The paper claims to provide a framework that converts unmodified ROS 2 applications into Lingua Franca (LF) programs for deterministic execution under logical time. It first enumerates a supported subset of ROS 2 features (publishers, subscribers, timers, and limited service/action patterns) that admit deterministic mapping, describes an automatic source-to-source transformation that emits LF reactors, and evaluates the result on a synthetic example plus the Autoware reference system, reporting that LF yields identical callback ordering and end-to-end latencies across runs while default ROS 2 exhibits nondeterministic ordering and variable latencies.

Significance. If the mapping and transformation preserve callback semantics for the supported subset, the work would allow existing ROS 2 codebases to obtain LF's logical-time guarantees, federated execution, and fault handling without source changes. The concrete evaluation on the Autoware reference system supplies direct evidence that the construction applies to a complex, production-grade system and produces the claimed determinism.

minor comments (2)
  1. The evaluation section reports only qualitative outcomes (deterministic order, consistent latencies) without quantitative details such as the number of runs performed, measured latency statistics, or explicit comparison tables; adding these would allow readers to assess the magnitude of improvement over default ROS 2.
  2. The description of the automatic conversion would benefit from a short table or diagram summarizing the exact ROS 2-to-LF reactor mapping for each supported feature (publishers, timers, etc.) to improve clarity of the transformation rules.

Simulated Author's Rebuttal

0 responses · 0 unresolved

We thank the referee for the positive summary, significance assessment, and recommendation of minor revision. No major comments were listed in the report, so we have no specific points requiring point-by-point rebuttal or changes at this stage. We remain available to incorporate any additional minor suggestions from the editor.

Circularity Check

0 steps flagged

No significant circularity identified

full rationale

The paper presents a source-to-source conversion framework that maps a supported subset of ROS 2 features (publishers, subscribers, timers, limited services/actions) onto Lingua Franca reactors whose logical-time semantics replace the nondeterministic ROS 2 executor. The derivation consists of an explicit feature enumeration followed by a concrete transformation rule set and direct empirical validation on both synthetic cases and the Autoware reference system; end-to-end outputs and callback ordering are shown to become deterministic under the LF schedule. No equations, fitted parameters, or predictions are involved, and no load-bearing claim reduces to a self-citation or self-definition. The construction is therefore self-contained against external benchmarks.

Axiom & Free-Parameter Ledger

0 free parameters · 1 axioms · 0 invented entities

The approach rests on domain assumptions about feature compatibility between ROS 2 and LF logical time; no free parameters or new entities are introduced.

axioms (1)
  • domain assumption Certain ROS 2 features can be executed deterministically under logical time without altering application semantics.
    This premise is stated explicitly as the foundation for identifying mappable features and building the automatic converter.

pith-pipeline@v0.9.1-grok · 5855 in / 1290 out tokens · 21438 ms · 2026-06-27T16:27:04.693499+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

33 extracted references · 1 linked inside Pith

  1. [1]

    Autoware reference system, 2026

    Autoware Foundation. Autoware reference system, 2026. https://github.com/ros-realtime/reference-system, Last Accessed: 2026-05-26

    2026

  2. [2]

    Bateni, M

    S. Bateni, M. Lohstroh, H. S. Wong, H. Kim, S. Lin, C. Menard, and E. A. Lee. Risk and mitigation of nondeterminism in distributed cyber-physical systems. In21st ACM/IEEE International Symposium on Formal Methods and Models for System Design (MEMOCODE), pages 1–11, 2023

    2023

  3. [3]

    Benveniste, P

    A. Benveniste, P. Caspi, S. Edwards, N. Halbwachs, P. Le Guernic, and R. de Simone. The synchronous languages 12 years later.Proceedings of the IEEE, 91(1):64–83, 2003

    2003

  4. [4]

    T. Blaß, D. Casini, S. Bozhko, and B. B. Brandenburg. A ROS 2 response-time analysis exploiting starvation freedom and execution-time variance. In2021 IEEE Real-Time Systems Symposium (RTSS), pages 41–53, 2021

    2021

  5. [5]

    T. Blaß, A. Hamann, R. Lange, D. Ziegenbein, and B. B. Brandenburg. Automatic latency management for ROS 2: Benefits, challenges, and open problems. In27th Real-Time and Embedded Technology and Applications Symposium (RTAS), pages 264–277. IEEE, 2021

    2021

  6. [6]

    Casini, T

    D. Casini, T. Blaß, I. Lütkebohle, and B. B. Brandenburg. Response- Time Analysis of ROS 2 Processing Chains Under Reservation-Based Scheduling. In S. Quinton, editor,31st Euromicro Conference on Real- Time Systems (ECRTS 2019), volume 133 ofLeibniz International Pro- ceedings in Informatics (LIPIcs), pages 6:1–6:23, Dagstuhl, Germany,

    2019

  7. [7]

    Schloss Dagstuhl – Leibniz-Zentrum für Informatik

  8. [8]

    Casini, J.-J

    D. Casini, J.-J. Chen, J. Li, F. Reghenzani, and H. Teper. A Survey of Real-Time Support, Analysis, and Advancements in ROS 2.Leibniz Transactions on Embedded Systems, 11(1):1:1–1:37, 2026

    2026

  9. [9]

    H. Choi, Y . Xiang, and H. Kim. Picas: New design of priority-driven chain-aware scheduling for ros2. In2021 IEEE 27th Real-Time and Embedded Technology and Applications Symposium (RTAS), pages 251–

  10. [10]

    Halbwachs, P

    N. Halbwachs, P. Caspi, P. Raymond, and D. Pilaud. The synchronous data flow programming language lustre.Proceedings of the IEEE, 79(9):1305–1320, 1991

    1991

  11. [11]

    W. K. Huang, Y . Terazawa, Y . Matsubara, and A. Iwai. Software-defined vehicles: Challenges and orchestrating mixed-criticality services using Lingua Franca.IEEE Embedded Systems Letters, 2025

    2025

  12. [12]

    S. Kato, S. Tokunaga, Y . Maruyama, S. Maeda, M. Hirabayashi, Y . Kit- sukawa, A. Monrroy, T. Ando, Y . Fujii, and T. Azumi. Autoware on board: Enabling autonomous vehicles with embedded systems. In2018 ACM/IEEE 9th International Conference on Cyber-Physical Systems (ICCPS), pages 287–296. IEEE, 2018

    2018

  13. [13]

    J. Kwok, S. Li, M. Lohstroh, and E. A. Lee. HPRM: High-performance robotic middleware for intelligent autonomous systems. InIEEE International Conference on Robotics and Automation (ICRA), pages 8093–8099, 2025

    2025

  14. [14]

    Lee and D

    E. Lee and D. Messerschmitt. Synchronous data flow.Proceedings of the IEEE, 75(9):1235–1245, 1987

    1987

  15. [15]

    Lee and T

    E. Lee and T. Parks. Dataflow process networks.Proceedings of the IEEE, 83(5):773–801, 05 1995

    1995

  16. [16]

    E. A. Lee, R. Akella, S. Bateni, S. Lin, M. Lohstroh, and C. Menard. Consistency vs. availability in distributed cyber-physical systems.ACM Trans. Embed. Comput. Syst., 22(5s):138:1–138:24, Sept. 2023

    2023

  17. [17]

    E. A. Lee, S. Bateni, S. Lin, M. Lohstroh, and C. Menard. Quantifying and generalizing the CAP theorem.arXiv, abs/2109.07771, 2021

    arXiv 2021

  18. [18]

    Lingua Franca coordination language, 2026

    Lingua Franca Project. Lingua Franca coordination language, 2026. https://github.com/lf-lang/lingua-franca, Version 0.12.1, Last Accessed: 2026-05-26

    2026

  19. [19]

    Playground for the Lingua Franca coordi- nation language, 2026

    Lingua Franca Project. Playground for the Lingua Franca coordi- nation language, 2026. https://github.com/lf-lang/playground-lingua- franca, Last Accessed: 2026-05-26

    2026

  20. [20]

    S. Liu, X. Jiang, N. Guan, Z. Wang, M. Yu, and W. Yi. Rtex: An efficient and timing-predictable multithreaded executor for ROS 2.IEEE Transactions on Computer-Aided Design of Integrated Circuits and Systems, 43(9):2578–2591, 2024

    2024

  21. [21]

    Lohstroh, C

    M. Lohstroh, C. Menard, S. Bateni, and E. A. Lee. Toward a lingua franca for deterministic concurrent systems.ACM Trans. Embed. Comput. Syst., 20(4):36:1–36:27, 2021

    2021

  22. [22]

    Lohstroh, Í

    M. Lohstroh, Í. Í. Romeo, A. Goens, P. Derler, J. Castrillón, E. A. Lee, and A. L. Sangiovanni-Vincentelli. Reactors: A deterministic model for composable reactive systems. In R. Chamberlain, M. Edin Grimheden, and W. Taha, editors,Cyber Physical Systems. Model-Based Design, pages 59–85, Cham, 2020. Springer International Publishing

    2020

  23. [23]

    Macenski, T

    S. Macenski, T. Foote, B. P. Gerkey, C. Lalancette, and W. Woodall. Robot Operating System 2: Design, architecture, and uses in the wild. Science Robotics, 7(66):eabm6074, 2022

    2022

  24. [24]

    Menard, M

    C. Menard, M. Lohstroh, S. Bateni, M. Chorlian, A. Deng, P. Donovan, C. Fournier, S. Lin, F. Suchert, T. Tanneberger, H. Kim, J. Castrillón, and E. A. Lee. High-performance deterministic concurrency using lingua franca.ACM Transactions on Architecture and Code Optimization, 20(4):48:1–48:29, 2023

    2023

  25. [25]

    Data distribution service (DDS) specifica- tion, version 1.4

    Object Management Group. Data distribution service (DDS) specifica- tion, version 1.4. Technical report, Object Management Group, 2015

    2015

  26. [26]

    ROS 2: Robot operating system 2, 2026

    Open Robotics. ROS 2: Robot operating system 2, 2026

    2026

  27. [27]

    Sagmeister, M

    S. Sagmeister, M. Weinmann, P. Pitschi, and M. Lienkamp. RSLCPP – deterministic simulations using ROS 2, 2026. https://arxiv.org/abs/2601.07052

    Pith/arXiv arXiv 2026

  28. [28]

    Y . Tang, N. Guan, X. Jiang, X. Luo, and W. Yi. Real-time performance analysis of processing systems on ros 2 executors. In2023 IEEE 29th Real-Time and Embedded Technology and Applications Symposium (RTAS), pages 80–92. IEEE, 2023

    2023

  29. [29]

    Y . Tang, X. Jiang, N. Guan, X. Luo, M. Yang, and W. Yi. Timing analysis of processing chains with data refreshing in ros 2.Journal of Systems Architecture, 155:103259, 2024

    2024

  30. [30]

    Teper, O

    H. Teper, O. Bell, M. Günzel, C. Gill, and J.-J. Chen. Reconciling ROS 2 with classical real-time scheduling of periodic tasks. In31st Real-Time and Embedded Technology and Applications Symposium (RTAS), pages 177–189. IEEE, 2025

    2025

  31. [31]

    Teper, T

    H. Teper, T. Betz, M. Günzel, D. Ebner, G. V on Der Brüggen, J. Betz, and J.-J. Chen. End-to-end timing analysis and optimization of multi- executor ROS 2 systems. In2024 IEEE 30th Real-Time and Embedded Technology and Applications Symposium (RTAS), pages 212–224, 2024

    2024

  32. [32]

    Teper, T

    H. Teper, T. Betz, G. V on Der Brüggen, K.-H. Chen, J. Betz, and J.-J. Chen. Timing-aware ROS 2 architecture and system optimization. In 29th International Conference on Embedded and Real-Time Computing Systems and Applications (RTCSA), pages 206–215. IEEE, 2023

    2023

  33. [33]

    Teper, M

    H. Teper, M. Günzel, N. Ueter, G. von der Brüggen, and J.-J. Chen. End-to-end timing analysis in ROS 2. In2022 IEEE Real-Time Systems Symposium (RTSS), pages 53–65, 2022. 12

    2022