Pith. sign in

REVIEW 4 major objections 4 minor 2 cited by

Circular Economy Design through System Dynamics Modeling

T0 review · 4 major / 4 minor · reviewed 2026-08-12 · deepseek-v4-flash

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

Pith's one-line read The paper proposes a quantitative definition of circularity, $\lambda(N) = -(m_{u,b} + \dot{m}_{u,c}\Delta)$, and claims that designing a circular system is the optimization problem $N^{*} = \arg\max \lambda(N)$, with Lagrange's equation…

desk verdict The paper's circularity metric is just negative mass balance, the analytical mechanics does no work, and the repair benefit is an artifact of an arbitrary time horizon. read the letter →

arxiv 2411.13540 v1 pith:52ODYF4M submitted 2024-11-20 math.DS

Federico Zocco , Monica Malvezzi This is my paper · ORCID
classification math.DS
keywords SDG12circulareconomycircularitymeasurethermodynamicalmaterialnetworkscompartmentaldynamicalsystemsanalyticalmechanicsrepairandreusememoryproperty
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

Circular economy thinking has lacked a dynamic, quantitative backbone; this paper supplies one by defining circularity $\lambda(N)$ as the negative of the total unsustainable mass leaving nonrenewable reservoirs or entering landfills, incinerators, or the environment. With that number, system design becomes an optimization statement: choose the compartmental network $N$ that maximizes $\lambda$, i.e., minimizes what leaves the loop. To compute $\lambda$, the paper models each network as a thermodynamic material network and uses Lagrange's equations, which it derives from the first law of thermodynamics, to track material motion. Three worked examples show reduction raises $\lambda$ (from $-2m$ to $-1.5m$ or $-1.6m$) and that repair plus reuse changes the time profile of $\lambda$: at the landfill time of the linear system the repaired system scores $-m$ instead of $-2m$, though both converge to $-2m$ once the repaired product also reaches landfill.

What carries the argument

A thermodynamical material network (TMN) is a set of connected thermodynamic compartments that transport, store, and transform a target material, represented compactly by a weighted mass-flow digraph: node-compartments carry mass stocks and arc-compartments carry mass flow rates. The circularity $\lambda$ in Eq. (2) is the score assigned to the whole network. Analytical mechanics enters through Proposition 1, which derives Lagrange's equation $\frac{d}{dt}\left(\frac{\partial L}{\partial \dot{s}}\right) = \frac{\partial L}{\partial s} + \xi$ from the first law of thermodynamics, so each transport or transformation step can be treated as a thermodynamic compartment and its dynamics computed. The memory property is encoded by evaluating $\lambda$ over a time horizon $\phi$, with all comparisons made at the same $\phi$.

What would settle it

Evaluate the linear and repair networks with identical mass $m$ and compute $\lambda_\phi$ for $\phi$ ranging from the first landfill arrival $t_{5l}$ to well past the second landfill arrival $t_{5p}$; because the paper's Example 3 already shows the ranking flips from repair-better to equal, the decisive test is whether the repair advantage survives when $\phi$ is set by a principled rule (e.g., expected product lifetime or a regulatory planning horizon) rather than chosen after the fact.

Watch

Extended reading notes

Core claim

The central claim is that circularity is a measurable, optimizable property of a networked system, not a vague ideal. For a thermodynamical material network $N$, circularity is $\lambda(N) = -(m_{u,b} + \dot{m}_{u,c}\Delta) \in (-\infty, 0]$, where $m_{u,b}$ is the total unsustainable batch mass and $\dot{m}_{u,c}$ is the total unsustainable continuous flow, with $\Delta$ an arbitrary but fixed conversion interval. The design task is $N^{*} = \arg\max \lambda(N)$. The paper's second claim is that $\lambda$ has memory: evaluated at a horizon $\phi$, the linear chain gives $\lambda_{\phi,l}|_{\phi=t_{5l}} = -2m$ while a repair-and-reuse network gives $\lambda_{\phi,p}|_{\phi=t_{5l}} = -m$; at the later horizon $t_{5p}$, both equal $-2m$. The repair configuration's advantage is therefore that material is kept in use for an additional time $\Delta_e = t_{5p} - t_{5l}$, which is captured only by the horizon-dependent $\lambda_\phi$.

Load-bearing premise

The score depends on two free choices a user must make, the conversion interval $\Delta$ and, once memory is included, the time horizon $\phi$; the paper gives no principled rule for choosing $\phi$, so the repair advantage it demonstrates could be an artifact of measuring at a favorable horizon rather than a property of the system.

Editorial extensions

If this is right

  • If $\lambda$ is accepted as the measure, circular design becomes the formal optimization $N^{*} = \arg\max \lambda(N)$: among candidate networks, choose the one with the largest circularity, i.e., the least unsustainable mass and flow.
  • Reducing material use is rewarded by this metric: using 50% renewable feedstock raises $\lambda$ from $-2m$ to $-1.5m$, and using 20% less material raises it to $-1.6m$.
  • Repair followed by reuse delays the landfill arrival, so at the linear system's landfill horizon the repaired system scores $-m$ versus $-2m$; both eventually reach $-2m$, making the repair benefit a time-window phenomenon.
  • Meaningful comparison of different networks requires holding $\Delta$ and the memory horizon $\phi$ fixed, because changing the horizon can change the ranking.
  • Robotic repair can be designed and evaluated as part of the network, with robot performance measured by its contribution to $\lambda$.

Reading between the lines

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

  • Because $\phi$ is a free choice, the ranking of repair against linear is not a system property until a principled horizon is fixed; a natural rule is expected product lifetime or a regulatory planning horizon, and the repair advantage should be tested across a range of such horizons.
  • The metric currently treats every unsustainable unit of mass as equal; weighting by material criticality, toxicity, or economic value would likely change which network is optimal.
  • Extending $\lambda$ to a dimensionless, weighted index for cross-plant comparison and adding energy or environmental-impact terms are direct next steps that the paper itself flags as future work.
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 / 4 minor

Summary. The paper proposes a quantitative circularity measure λ(N) based on compartmental mass accounting, defines the design of circular systems as an arg-max problem over λ(N), and claims to apply analytical mechanics (Lagrange's equations) to circularity. Three worked examples are given: a linear take-make-dispose chain, two material-reduction cases, and a repair-and-reuse network. The repair example introduces a 'memory property' of λ, where the circularity is evaluated with respect to a time horizon φ, and the paper reports that at φ = t5l repair outperforms the linear system. The paper then suggests robotic repair as a future application area.

Significance. If the proposed measure were sound, a quantitative, dynamics-based circularity indicator would be a useful complement to material flow analysis. The paper does provide simple, correctly evaluated batch-mass examples (Eqs. (18)-(20)) and a clear statement of the design objective (3). However, the central contributions are not supported: the analytical-mechanics equations are not used to compute any of the reported circularity values, and the 'memory property' is an artifact of an arbitrarily chosen time horizon. The manuscript's novelty reduces to mass counting with a truncated time window, and the claimed connection to analytical mechanics is not demonstrated.

major comments (4)
  1. [§3.3, Eqs. (2)-(3)] The memory property is introduced informally in Example 3 without a formal definition of λ_φ. The values λ_{φ,l}|_{φ=t5l} = -2m, λ_{φ,p}|_{φ=t5l} = -m, and λ_{φ,l}|_{φ=t5p} = λ_{φ,p}|_{φ=t5p} = -2m are asserted by counting whether the mass has crossed the landfill boundary by the horizon φ. Because φ is an arbitrary user-chosen constant and no principled rule for selecting it is provided, the design problem (3) is ill-posed: two analysts using different horizons will select different optimal networks. The ranking of the linear vs repair network reverses with φ, so the claim that the memory property 'highlights the benefit of repair' is an artifact of the measurement window, not an intrinsic property of the repair network.
  2. [§3.1 and §3.3, Eqs. (8)-(17)] The analytical-mechanics equations (8)-(17) are never used to compute the circularity values or the landfill times t5l and t5p. The reported results λ_l = -2m, λ_p = -2m, and λ_{φ,p}|_{φ=t5l} = -m follow solely from mass accounting under Definition 3. The ordering t4 < t5l << t5p is assumed, not derived from the equations of motion. Consequently, the central claim that analytical mechanics is applied to circularity is not supported by the examples; the equations are a free-standing exposition that does no work in the analysis.
  3. [§3.3] The text first states that 'Equation (2) yields that the circularity with the repair stage is equivalent to the linear case... λp = λl = -2m' and then immediately introduces λ_{φ,p}|_{φ=t5l} = -m. If λ_φ is intended to be a horizon-truncated version of (2), the original Definition 4 needs to be generalized in Section 2; if λ_φ is a different measure, the change is unannounced. As written, the relationship between λ(N) and λ_φ(N) is ambiguous, making the example internally inconsistent.
  4. [Definition 4, §3.3] The justification for the arbitrary conversion interval Δ ('without loss of generality, Δ = 1 s') does not extend to the horizon φ. Δ is a unit-conversion constant with the same value in all comparisons, whereas φ is a physical time that changes which events are counted. The paper asserts that 'comparisons must be done for the same memory φ', but it offers no rule for choosing φ, and unlike Δ, different choices of φ produce different qualitative rankings, not merely different numerical scales.
minor comments (4)
  1. [§3.1, Eqs. (12) and (15)] Equations (12) and (15) mix boundary conditions with time-interval statements (e.g., '¨s(t1) = ˙s(t1) = 0, s(t1) = l5, t1 < t < t2'); these should be rewritten as separate conditions for clearance.
  2. [§3.2] In the first reduction case, the paper assumes that the renewable fraction also enters the landfill, so the total landfill mass is m. This modeling choice should be stated explicitly, since it affects the comparison between λ_r1 and λ_r2.
  3. [§3.3] The phrase 'same circularity for t > t5p' is confusing because φ is a fixed time horizon, not a running variable; the sentence should say 'for φ > t5p'.
  4. [§2, Proposition 1] The proof of Proposition 1 is entirely delegated to reference [14]. Since the analytical-mechanics connection is a stated contribution, a short indication of the derivation (or at least a statement of the assumptions) would strengthen the paper.

Circularity Check

1 steps flagged · score 8.0 of 10

The claimed repair benefit is horizon-dependent: λφ counts landfill arrivals up to φ, so choosing φ=t5l forces λφ,p=−m, while at φ=t5p the advantage vanishes; the central 'memory' result is self-definitional.

  1. self definitional [Section 3.3, Example 3 (Case of Robotic Repair), third paragraph, after Eq. (2) and Definition 4]
    "This phenomenon shows that circularity has a memory, that is, it must be defined with respect to a time horizon, which we call ϕ. ... Calculating the circularity with respect to ϕ = t5l yields λϕ,l|ϕ=t5l = −2m and λϕ,p|ϕ=t5l = −m ... which highlights the benefit of repair. In contrast, if ϕ = t5p , we have that λϕ,l|ϕ=t5p = λϕ,p|ϕ=t5p = −2m"

    The paper never formally defines λφ or gives an equation for it; the displayed values are obtained by applying the accounting rule already contained in Eq. (2) only up to the chosen horizon φ. At φ=t5l the linear mass has reached the landfill and is counted as an extra −m, while the repaired mass has not yet arrived, so its landfill contribution is omitted, forcing λφ,p=−m. At φ=t5p both masses have arrived, forcing equality. Thus the 'memory property' and the 'benefit of repair' are not derived from the dynamics or from Eq. (2); they are put into the measure by the arbitrary choice of horizon. Since no principled rule for choosing φ is given, the design objective (3) is ill-posed: different analysts choosing different horizons obtain different rankings of the same networks.

full rationale

The basic circularity counts are self-contained: λl=−2m, λr1=−1.5m, and λr2=−1.6m follow directly from Definition 3 by mass counting, so those parts are not circular. The analytical-mechanics equations (8)–(16) are not used to compute any λ value, and Proposition 1's citation to prior work [14] is therefore not load-bearing for the main claim. The central claimed discovery, however, is the 'memory property' showing repair improves circularity. That result is not a consequence of Definition 4: the paper informally extends λ(N) to λφ without a formula, and then counts whether the mass has reached the landfill within φ. Choosing φ=t5l before the repaired mass arrives forces λφ,p=−m, while choosing φ=t5p makes the two cases equal. The paper even admits comparisons depend on keeping φ fixed, but gives no criterion for selecting φ, so the ranking and the arg-max objective (3) are determined by an arbitrary input rather than by the system. This is self-definitional circularity: the advertised benefit of repair is equivalent to the chosen measurement horizon.

Assumptions & free parameters 2 free parameters · 5 assumptions · 1 invented entities

The ledger shows the measure λ rests on two hand-picked constants (Δ and φ) and on assumptions from the authors' prior work. The only invented entity is λ itself, which has no independent falsifiable handle.

free parameters (2)
  • Δ (time conversion interval) = 1 s (assumed)
    Introduced in Definition 4 to convert the continuous unsustainable flow mdot_u,c into a mass. The paper states 'The choice of Δ is arbitrary, but its value must be kept the same for any calculation'. It is a hand-chosen constant.
  • φ (memory time horizon) = t5l and t5p in Example 3
    The circularity with memory λφ depends on a time horizon φ chosen for comparison. The ranking of linear vs. repair networks reverses depending on this choice, so φ functions as a free parameter.
assumptions (5)
  • domain assumption Proposition 1: Lagrange's equation of motion can be derived from the first law of thermodynamics (proof cited to [14])
    This theorem from the authors' prior work is used to justify representing mechanical systems as thermodynamic compartments; the proof is not reproduced.
  • domain assumption A mechanical system can, in general, be represented as a thermodynamic compartment
    Inferred from Proposition 1 at the end of Section 2; motivates the examples but is not used in computing λ.
  • ad hoc to paper Unsustainable mass is the only quantity needed to assess circularity
    Definition 3-4 equates circularity with the negative total unsustainable mass, ignoring material quality, recyclability, energy, and time beyond the chosen horizon.
  • ad hoc to paper A fixed interval Δ can convert continuous flow to mass without affecting comparisons
    Definition 4; 'without loss of generality, we assume Δ=1s'. This is a dimensional convention, not derived from physics.
  • ad hoc to paper Meaningful comparisons require the same memory horizon φ
    Section 3.3: comparisons must be done for the same φ; no principle for selecting φ is given.
invented entities (1)
  • λ(N), the circularity measure
    purpose: Quantify circularity of a thermodynamical material network and serve as objective in arg max design
    The scalar index is defined by the authors with no external benchmark or falsifiable prediction; its value and ranking depend on the arbitrary parameters Δ and φ.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Circular Economy Design through System Dynamics Modeling." pith.science (2026). https://pith.science/paper/52ODYF4M

@misc{pith2026241113540,
  author       = {Pith},
  title        = {Pith review of: Circular Economy Design through System Dynamics Modeling},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/52ODYF4M}},
  note         = {Machine review of arXiv:2411.13540}
}
abstract

Nowadays, there is an increasing concern about the unsustainability of the take-make-dispose paradigm upon which traditional production and consumption systems are built. The concept of circular economy is gaining attention as a potential solution, but it is an emerging field still lacking analytical and methodological dynamics approaches. Hence, in this paper, firstly we propose a quantitative definition of circularity, namely, $\lambda$, predicated on compartmental dynamical thermodynamics, and then, we use it to state the optimization of the circularity $\lambda$ as an arg-max problem. By leveraging the derivation of Lagrange's equations of motion from the first law of thermodynamics, we apply the analytical mechanics approaches to circularity. Three examples illustrate the calculation of $\lambda$ for different settings of two compartmental networks. In particular, hypothesizing a repair stage followed by product reuse we highlight the memory property of $\lambda$. Finally, robotic repair is proposed within this framework to pave the way for circular robotics as a new area of research in which the performance of a robotic system is measured against $\lambda$.

Figures

Figures reproduced from arXiv: 2411.13540 by the authors.

Figure 1
Figure 1. Classic (left) and generalized (right) representations of the Rankine cycle. Fig￾ure adapted from [13]. of [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗
Figure 2
Figure 2. Geometric diagraph of Examples 1, 2, and 3. is extracted from nonrenewable reservoirs; and second, all the material is sent to a landfill after its first use. The dynamics of the cube of mass m and material β can be described by considering one compartment at a time starting from the nonrenewable reservoir (c 1 1,1 ). Assume that the cube leaves the reservoir in t0 = 0. Let Ft be the force of the truck (c 5 1,2 ), l… view at source ↗

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 2 Pith papers

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

  1. CIRO7.2: A Material Network with Circularity of -7.2 and Reinforcement-Learning-Controlled Robotic Disassembler

    cs.RO 2025-06 conditional novelty 4.0 of 10

    The circularity values reported (-2.1 to -7.2) are direct evaluations of the authors' own metric, and the headline sensitivity result is an algebraic consequence of that metric, not an empirical finding.

  2. Circular Microalgae-Based Carbon Control for Net Zero

    math.DS 2025-02 reject novelty 4.0 of 10

    The paper derives a 625:1 microalgae-to-digester volume requirement for net zero and uses RL to raise modeled CO2 uptake, but the volume ratio rests on incompatible units.

Reference graph

Works this paper leans on

14 extracted references · 12 canonical work pages · cited by 2 Pith papers

  1. [1]

    Global Journal of Flexible Systems Management 25(3), 419–443 (2024)

    Basile, V., Petacca, N., Vona, R.: Measuring circularity in life cycle management: A literature review. Global Journal of Flexible Systems Management 25(3), 419–443 (2024)

    work page 2024

  2. [2]

    Bondy, J.A., Murty, U.S.R.: Graph theory with applications, vol. 290. Macmillan London (1976)

    work page 1976

  3. [3]

    Annual Review of Ma- terials Research 52(1), 525–559 (2022)

    Cullen, J.M., Cooper, D.R.: Material flows and efficiency. Annual Review of Ma- terials Research 52(1), 525–559 (2022)

    work page 2022

  4. [4]

    Ellen MacArthur Foundation: What is a circular economy?, webpage: https://www.ellenmacarthurfoundation.org/topics/ circular-economy-introduction/overview ; last access: 16 October 2024

    work page 2024

  5. [5]

    ec.europa.eu/topics/waste-and-recycling_en ; last access: 16 October 2024

    European Commission: Waste and recycling, webpage: https://environment. ec.europa.eu/topics/waste-and-recycling_en ; last access: 16 October 2024

    work page 2024

  6. [6]

    European Commission: Critical raw materials (2023), webpage: https://single-market-economy.ec.europa.eu/sectors/raw-materials/ areas-specific-interest/critical-raw-materials_en ; last access: 16 October 2024

    work page 2023

  7. [7]

    Entropy 19(11), 621 (2017)

    Haddad, W.M.: Thermodynamics: The unique universal science. Entropy 19(11), 621 (2017)

    work page 2017

  8. [8]

    Princeton Uni- versity Press (2019)

    Haddad, W.M.: A dynamical systems theory of thermodynamics. Princeton Uni- versity Press (2019)

    work page 2019

Show all 14 references
  1. [9]

    John Wiley & Sons (2010)

    Moran, M.J., Shapiro, H.N., Boettner, D.D., Bailey, M.B.: Fundamentals of engi- neering thermodynamics. John Wiley & Sons (2010)

    2010

  2. [10]

    Planbureau voor de Leefomgeving (issue 2544, report) (2017)

    Potting, J., Hekkert, M.P., Worrell, E., Hanemaaijer, A.: Circular economy: Mea- suring innovation in the product chain. Planbureau voor de Leefomgeving (issue 2544, report) (2017)

    2017

  3. [11]

    Journal of Industrial Ecology (2024)

    Wang, J., Ray, K., Brito-Parada, P., Plancherel, Y., Bide, T., Mankelow, J., Morley, J., Stegemann, J.A., Myers, R.: Bayesian material flow analysis for systems with multiple levels of disaggregation and high dimensional data. Journal of Industrial Ecology (2024)

    2024

  4. [12]

    arXiv preprint arXiv:2209.15051 (2024)

    Zocco, F.: Circularity of thermodynamical material networks: Indicators, examples, and algorithms. arXiv preprint arXiv:2209.15051 (2024)

    2024

  5. [13]

    IEEE Access pp

    Zocco, F., Haddad, W.M., Corti, A., Malvezzi, M.: A unification be- tween deep-learning vision, compartmental dynamical thermodynamics, and robotic manipulation for a circular economy. IEEE Access pp. 1–1 (2024). https://doi.org/10.1109/ACCESS.2024.3493755

    2024

  6. [14]

    Interna- tional Journal of Sustainable Engineering 16(1), 1–14 (2023)

    Zocco, F., Sopasakis, P., Smyth, B., Haddad, W.M.: Thermodynamical material networks for modeling, planning, and control of circular material flows. Interna- tional Journal of Sustainable Engineering 16(1), 1–14 (2023)

    2023

Pith tools

Reviewed August 12, 2026 · model on record in the stance chip above.