pith. sign in

arxiv: 2604.17417 · v1 · submitted 2026-04-19 · 💻 cs.SE · cs.AI· cs.SY· eess.SY

Project resilience as network robustness

Sebastiano A. Piccolo , Giorgio Terracina This is my paper

Pith reviewed 2026-05-10 05:38 UTC · model grok-4.3

classification 💻 cs.SE cs.AIcs.SYeess.SY
keywords project resiliencenetwork robustnesspersonnel lossvulnerability assessmenttask dependency networksengineering project managementkey person risk
0
0 comments X

The pith

Modeling a project as a network and measuring its robustness to node removal produces more consistent estimates of resilience to losing key personnel than existing methods.

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

Engineering projects rely on uneven labor distribution, where some people handle many tasks and serve as integration points. Existing risk estimates either assume the best case or overlook how losing central people fragments the remaining work. The paper reframes the project as a network of personnel and task dependencies, then applies robustness metrics to quantify how much function survives when central nodes disappear. This yields estimates that track fragmentation more closely and remain comparable across projects. If the reframing holds, it supplies a practical way to flag real disruption risks before they cause delays or added costs.

Core claim

By representing the project as a network in which nodes stand for team members and edges capture task dependencies and integration roles, robustness metrics can be applied directly to the loss of high-centrality nodes; the resulting scores avoid both the optimism of best-case assumptions and the bias introduced when fragmentation is ignored, delivering more consistent vulnerability rankings than prior techniques.

What carries the argument

The project network whose nodes are personnel and whose edges encode task assignments and integration responsibilities; robustness is measured by the persistence of connectivity or task coverage after successive removal of the most central nodes.

If this is right

  • Vulnerability rankings become directly comparable across different projects and teams.
  • Knowledge-transfer and retention efforts can be prioritized toward nodes whose removal produces the largest measured drop in network function.
  • Multiple simultaneous departure scenarios can be simulated within a single consistent framework.
  • Project designs that spread integration roles can be evaluated by their effect on the same robustness metric.

Where Pith is reading between the lines

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

  • The same network construction could be applied to open-source or volunteer collaborations where task graphs are already public.
  • Longitudinal tracking of how edges change over time would test whether static snapshots suffice or whether dynamic updates improve prediction.
  • Organizations could combine the robustness score with external data on replacement cost to produce a single risk-dollar figure.

Load-bearing premise

That the static network of assigned tasks and reported dependencies accurately reflects the actual flow of work and the real consequences of losing any given person.

What would settle it

An empirical study that records actual project delays, rework, or cost overruns after documented departures of central personnel and checks whether the network-derived robustness scores predict those outcomes more accurately than the optimistic or fragmentation-blind alternatives.

Figures

Figures reproduced from arXiv: 2604.17417 by Giorgio Terracina, Sebastiano A. Piccolo.

Figure 1
Figure 1. Figure 1: Comparison of the Bus Factor measures formalized as combinatorial problems on bipartite graphs. Unfortunately, as we show, the available methods to estimate Bus Factor suffer from severe short￾comings: 1) they are often not generally applicable, being tailored to specific metadata (e.g., GitHub); 2) they depend on arbitrary thresholds to define stalling conditions; 3) they fail to capture the resulting pro… view at source ↗
Figure 2
Figure 2. Figure 2: Sensitivity of Bus Factor measures to changes in network density (RQ1). A) Densification: MCS and ℬ𝐵𝐹 increase with network density, as expected. Inset: values of MCS and ℬ𝐵𝐹 expressed as number of people. B) Sparsification: MCS and ℬ𝐵𝐹 decrease with network density, as expected. Inset: a zoomed-in view of the range [5000, 10000] edges removed, showing that ℬ𝐵𝐹 is more stable than MCS. Approximation. Since… view at source ↗
Figure 3
Figure 3. Figure 3: Sensitivity of Bus Factor measures to personnel redundancy (Q2). A) Sensitivity to singletons. B) Sensitivity to duplicates. In the inset: values of MCS and ℬ𝐵𝐹 expressed as number of people. Because ℬ𝐵𝐹 measures connected components rather than simple coverage, it accurately reflects the gradual degradation or improvement of the project structural integrity without threshold-induced instability. Sensitivi… view at source ↗
Figure 4
Figure 4. Figure 4: Comparison of the Bus Factor of a real-world project and its version optimized for robustness, against that of a statistical null model. A: decay curves. B: permutation test. Optimization via Simulated Annealing. To correct this structural vulnerability, we treat the task assignment as an optimization problem. We employ a Simulated Annealing (SA) to iteratively rewire the bipartite graph in order to maximi… view at source ↗
read the original abstract

Engineering projects are the result of the combined effort of their members. Yet, it has been documented that labor division withing projects is unevenly distributed: some project members are specialists undertaking only few tasks, whereas other are generalists and are responsible for the success of many tasks. Moreover, the latter are often facilitators of project integration. Such a workload distribution prompts one question: how resilient is a project to key personnel loss? Far from being a theoretical problem, the reliance of a project on a few key people can lead to severe economic losses and delays. We argue that current methods to estimate such a risk are unsatisfactory: some methods offer a best-case estimate and are, therefore, too optimistic; other methods fail to capture project fragmentation leading to biased estimates and unrealistic consequences in many settings. In this paper, we develop a novel method to assess project vulnerability by looking at it from the lens of network robustness. We compare our method against existing alternatives and show that it offers better and more consistent estimates of project resilience to personnel loss.

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

2 major / 1 minor

Summary. The manuscript proposes modeling engineering projects as networks to assess resilience to key personnel loss via robustness metrics. It critiques prior methods as either overly optimistic (best-case estimates) or biased (failing to capture fragmentation), introduces a novel network-based procedure, and claims via comparisons that the new approach yields better and more consistent resilience estimates.

Significance. If the network model were shown to align with observed project outcomes, the method could provide project managers with a structured way to quantify vulnerability to staff departures. However, the lack of any grounding in real disruptions means the practical significance cannot be assessed from the current manuscript.

major comments (2)
  1. [Abstract] Abstract: the central claim that the method 'offers better and more consistent estimates of project resilience' is asserted without any data, metrics, comparison protocol, or results. This assertion is load-bearing for the paper's contribution and must be supported with concrete evidence.
  2. [Method] The modeling assumption that network robustness metrics accurately reflect real-world integration roles and consequences of personnel loss is unvalidated; no comparison to documented delays, cost overruns, or other observed outcomes after key departures is provided. This directly affects whether the reported superiority is externally meaningful.
minor comments (1)
  1. [Abstract] Abstract: 'labor division withing projects' contains a typo and should read 'within'.

Simulated Author's Rebuttal

2 responses · 1 unresolved

We thank the referee for the constructive feedback. We agree that the abstract requires more specificity and that external validation against real outcomes would strengthen the work, though our contribution is primarily methodological. We address each comment below and have made revisions accordingly.

read point-by-point responses

  1. Referee: [Abstract] Abstract: the central claim that the method 'offers better and more consistent estimates of project resilience' is asserted without any data, metrics, comparison protocol, or results. This assertion is load-bearing for the paper's contribution and must be supported with concrete evidence.

    Authors: The manuscript body details the comparison protocol on synthetic networks derived from task dependency structures, using metrics such as variance in estimated resilience scores and consistency across multiple personnel removal scenarios. The abstract, being a high-level summary, omitted these specifics. We will revise the abstract to briefly reference the comparison metrics and protocol while preserving its length. revision: yes

  2. Referee: [Method] The modeling assumption that network robustness metrics accurately reflect real-world integration roles and consequences of personnel loss is unvalidated; no comparison to documented delays, cost overruns, or other observed outcomes after key departures is provided. This directly affects whether the reported superiority is externally meaningful.

    Authors: We acknowledge the assumption that network robustness proxies integration roles and that superiority is shown relative to prior methods' internal biases (optimism and fragmentation) rather than predictive accuracy on real disruptions. The study uses controlled synthetic experiments to demonstrate consistency advantages. We will add an explicit limitations section noting the absence of real-world outcome data and suggesting it as future work. revision: partial

standing simulated objections not resolved
  • Direct empirical comparison against documented real-world project delays or cost overruns following key personnel losses, as no such observational dataset is incorporated in the current study.

Circularity Check

0 steps flagged

No significant circularity detected in derivation chain

full rationale

The paper introduces a network-robustness method for project resilience as a novel alternative to existing approaches, which it critiques as optimistic or biased. No self-definitional reductions, fitted inputs renamed as predictions, or load-bearing self-citations appear in the provided abstract or claims. The core contribution is a comparative evaluation of the new method against alternatives, presented without equations or steps that reduce by construction to the inputs themselves. The derivation remains self-contained against external benchmarks.

Axiom & Free-Parameter Ledger

0 free parameters · 0 axioms · 0 invented entities

No free parameters, axioms, or invented entities are mentioned in the abstract; the method is described at a high level without technical details.

pith-pipeline@v0.9.0 · 5477 in / 1043 out tokens · 48231 ms · 2026-05-10T05:38:38.139661+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

22 extracted references · 22 canonical work pages

  1. [1]

    S. A. Piccolo, S. Lehmann, A. Maier, Design process robustness: a bipartite network analysis reveals the central importance of people, Design Science 4 (2018) e1

    work page 2018

  2. [2]

    S. A. Piccolo, P. De Meo, G. Terracina, Evaluating and improving projects’ bus-factor: a network analytical framework, in: 16th International Conference, ASONAM 2024, Rende, Italy, September 2–5, 2024, Proceedings, Part I, LNCS, Springer, 2024

    work page 2024

  3. [3]

    S. A. Piccolo, P. De Meo, G. Terracina, G. Greco, The theory and practice of computing the bus factor, Information Sciences 743 (2026) 123353. URL: https://www.sciencedirect.com/science/ article/pii/S0020025526002847. doi:https://doi.org/10.1016/j.ins.2026.123353

    work page doi:10.1016/j.ins.2026.123353 2026

  4. [4]

    In: Proceedings of the 14th International Workshop on Principles of Software Evolution

    K. Yamashita, S. McIntosh, Y. Kamei, A. E. Hassan, N. Ubayashi, Revisiting the applicability of the pareto principle to core development teams in open source software projects, in: Proceedings of the 14th International Workshop on Principles of Software Evolution, IWPSE 2015, Association for Computing Machinery, New York, NY, USA, 2015, p. 46–55. URL: htt...

    work page doi:10.1145/2804360 2015

  5. [5]

    Agrawal, A

    A. Agrawal, A. Rahman, R. Krishna, A. Sobran, T. Menzies, We don’t need another hero? the impact of "heroes" on software development, in: Proceedings of the 40th International Conference on Software Engineering: Software Engineering in Practice, ICSE-SEIP ’18, Association for Computing Machinery, New York, NY, USA, 2018, p. 245–253. URL: https://doi.org/1...

    work page doi:10.1145/3183519.3183549 2018

  6. [6]

    Majumder, J

    S. Majumder, J. Chakraborty, A. Agrawal, T. Menzies, Why software projects need heroes (lessons learned from 1100+ projects), arXiv preprint arXiv:1904.09954 (2019)

    work page arXiv 1904

  7. [7]

    S. A. Piccolo, A. M. Maier, S. Lehmann, C. A. McMahon, Iterations as the result of social and technical factors: empirical evidence from a large-scale design project, Research in Engineering Design 30 (2019) 251–270

    work page 2019

  8. [8]

    M. S. Zanetti, I. Scholtes, C. J. Tessone, F. Schweitzer, The rise and fall of a central contributor: Dynamics of social organization and performance in the gentoo community, in: 2013 6th interna- tional workshop on cooperative and human aspects of software engineering (chase), IEEE, 2013, pp. 49–56

    work page 2013

  9. [9]

    Russo Latona, C

    G. Russo Latona, C. Gote, C. Zingg, G. Casiraghi, L. Verginer, F. Schweitzer, Shock! quantifying the impact of core developers’ dropout on the productivity of oss projects, in: Companion Proceedings of the ACM Web Conference 2024, WWW ’24, Association for Computing Machinery, New York, NY, USA, 2024, p. 706–709. URL: https://doi.org/10.1145/3589335.365155...

    work page doi:10.1145/3589335.3651559 2024

  10. [10]

    Zazworka, K

    N. Zazworka, K. Stapel, E. Knauss, F. Shull, V. R. Basili, K. Schneider, Are developers complying with the process: an xp study, in: Proceedings of the 2010 ACM-IEEE International Symposium on Empirical Software Engineering and Measurement, ESEM ’10, Association for Computing Machinery, New York, NY, USA, 2010. URL: https://doi.org/10.1145/1852786.1852805...

    work page doi:10.1145/1852786.1852805 2010

  11. [11]

    Cosentino, J

    V. Cosentino, J. L. C. Izquierdo, J. Cabot, Assessing the bus factor of git repositories, in: 2015 IEEE 22nd International Conference on Software Analysis, Evolution, and Reengineering (SANER), 2015, pp. 499–503. doi:10.1109/SANER.2015.7081864

    work page doi:10.1109/saner.2015.7081864 2015

  12. [12]

    4 Guilherme Avelino, Leonardo Passos, Andre Hora, and Marco Tulio Valente

    G. Avelino, L. Passos, A. Hora, M. T. Valente, A novel approach for estimating truck factors, in: 2016 IEEE 24th International Conference on Program Comprehension (ICPC), 2016, pp. 1–10. doi:10.1109/ICPC.2016.7503718

    work page doi:10.1109/icpc.2016.7503718 2016

  13. [13]

    Ferreira, L

    F. Ferreira, L. L. Silva, M. T. Valente, Turnover in open-source projects: The case of core developers, in: Proceedings of the XXXIV Brazilian Symposium on Software Engineering, 2020, pp. 447–456

    work page 2020

  14. [14]

    Foucault, M

    M. Foucault, M. Palyart, X. Blanc, G. C. Murphy, J.-R. Falleri, Impact of developer turnover on quality in open-source software, in: Proceedings of the 2015 10th joint meeting on foundations of software engineering, 2015, pp. 829–841

    work page 2015

  15. [15]

    Avelino, E

    G. Avelino, E. Constantinou, M. T. Valente, A. Serebrenik, On the abandonment and survival of open source projects: An empirical investigation, in: 2019 ACM/IEEE International Symposium on Empirical Software Engineering and Measurement (ESEM), IEEE, 2019, pp. 1–12

    work page 2019

  16. [16]

    InProceedings of the 44th International Conference on Software Engineering: Software Engineering in Practice (ICSE-SEIP ’22)

    E. Jabrayilzade, M. Evtikhiev, E. Tüzün, V. Kovalenko, Bus factor in practice, in: Proceedings of the 44th International Conference on Software Engineering: Software Engineering in Practice, ICSE-SEIP ’22, Association for Computing Machinery, New York, NY, USA, 2022, p. 97–106. URL: https://doi.org/10.1145/3510457.3513082. doi:10.1145/3510457.3513082

    work page doi:10.1145/3510457.3513082 2022

  17. [17]

    Ricca, A

    F. Ricca, A. Marchetto, M. Torchiano, On the difficulty of computing the truck factor, in: Proceed- ings of the 12th International Conference on Product-Focused Software Process Improvement, PROFES’11, Springer-Verlag, Berlin, Heidelberg, 2011, p. 337–351

    work page 2011

  18. [18]

    Ferreira, M

    M. Ferreira, M. T. Valente, K. Ferreira, A comparison of three algorithms for computing truck factors, in: 2017 IEEE/ACM 25th International Conference on Program Comprehension (ICPC), 2017, pp. 207–217. doi:10.1109/ICPC.2017.35

    work page doi:10.1109/icpc.2017.35 2017

  19. [19]

    M. R. Garey, D. S. Johnson, Computers and intractability, volume 174, freeman San Francisco, 1979

    work page 1979

  20. [20]

    C. M. Schneider, A. A. Moreira, J. S. Andrade Jr, S. Havlin, H. J. Herrmann, Mitigation of malicious attacks on networks, Proceedings of the National Academy of Sciences 108 (2011) 3838–3841

    work page 2011

  21. [21]

    P. R. Lawrence, J. W. Lorsch, Organization and environment: Managing differentiation and inte- gration, Harvard Business Review Press, 1986

    work page 1986

  22. [22]

    Heath, N

    C. Heath, N. Staudenmayer, Coordination neglect: How lay theories of organizing complicate coordination in organizations, Research in organizational behavior 22 (2000) 153–191

    work page 2000