0 cumulative citations
View corpus contextManufacturers that put AI at the center of supply-chain decision-making cut waste by up to 26% and boost material recovery by as much as 17%; mere digitalization without AI orchestration delivers no measurable circular benefits.
AI-Driven Circular Digital Supply Chains: An Integrated Framework for Sustainable Value Creation in Emerging Markets
openalex
quasi_experimental
medium evidence
7/10 relevance
Summary only
summary available; pdf_status=paywall
DOI
Source
PDF
In eight manufacturing firms in Mexico and Colombia, adopting AI-centric decision architectures reduced waste by 18–26% and increased material reuse/recovery by 14–17%, whereas firms that only digitized without AI saw no significant circular improvements.
Citation observations
Cumulative provider counts captured on specific dates; providers are never combined.
Digital transformation has improved visibility and efficiency in supply chains, yet it has delivered limited progress toward circular economy objectives, particularly in emerging markets. Existing research has largely examined digital technologies in isolation or treated artificial intelligence (AI) as a secondary analytical tool, leaving unclear whether circular performance improvements stem from technology adoption itself or from changes in supply chain decision-making. This study addresses this gap by proposing the AI-Driven Circular Digital Supply Chain (AICD-SC) framework, which conceptualizes AI as a central decision orchestrator integrating predictive, prescriptive, and simulation-based capabilities to coordinate closed-loop supply chain processes. The study employs a sequential explanatory mixed-methods design, combining 17 semi-structured interviews with a quasi-experimental Difference-in-Differences analysis of eight manufacturing firms in Mexico and Colombia during 2023–2024. The results show that firms adopting AI-centric decision architectures achieved waste reductions of 18–26% and improvements in material reuse and recovery of 14–17%, while firms relying on digitally enabled but non–AI-centric configurations exhibited no statistically significant circular performance gains. These findings indicate that circular outcomes do not emerge from digitalization alone, but from how supply chain decisions are architected and orchestrated through AI. The study concludes by offering a phased adoption roadmap aligned with Sustainable Development Goal 12, providing actionable implications for managers and policymakers in emerging markets.
Summary
Main Finding
AI-centric decision architectures — where AI acts as a central orchestrator (integrating predictive, prescriptive and simulation capabilities) — produce meaningful circular economy gains in manufacturing supply chains in emerging markets. In the study sample, firms that adopted such AI-driven architectures achieved waste reductions of 18–26% and material reuse/recovery improvements of 14–17%. Firms that adopted digital technologies without an AI-centric decision architecture showed no statistically significant circular performance gains. The implication: digitalization alone is insufficient; the way supply-chain decisions are architected and orchestrated via AI determines circular outcomes.
Key Points
- Research gap addressed: prior work treated digital technologies in isolation or used AI only as an analytical tool; this study isolates AI as the decision orchestrator and tests its effect on circular performance.
- Conceptual contribution: introduces the AI-Driven Circular Digital Supply Chain (AICD-SC) framework — positions AI at the center of closed-loop coordination by combining predictive (forecasting), prescriptive (optimization/policy), and simulation (scenario testing) capabilities.
- Empirical finding: AI-centric firms realized 18–26% reductions in waste and 14–17% increases in material reuse/recovery over 2023–2024.
- Comparative finding: digitally enabled but non–AI-centric configurations yielded no statistically significant improvements in circular metrics.
- Context: study focused on eight manufacturing firms in Mexico and Colombia (2023–2024) and included qualitative validation through 17 semi-structured interviews.
- Practical output: the paper provides a phased adoption roadmap aligned with SDG 12 (responsible consumption and production), aimed at managers and policymakers in emerging markets.
Data & Methods
- Research design: sequential explanatory mixed-methods.
- Qualitative: 17 semi-structured interviews to develop and validate the AICD-SC framework, understand implementation pathways, barriers, and organizational changes.
- Quantitative: quasi-experimental Difference-in-Differences (DiD) analysis on eight manufacturing firms in Mexico and Colombia over 2023–2024 to estimate causal effects of adopting AI-centric decision architectures on circular outcomes.
- Outcomes measured: waste generation (reduction) and material reuse/recovery rates (improvement).
- Key empirical results: statistically significant reductions in waste (18–26%) and increases in reuse/recovery (14–17%) for AI-centric adopters; no significant changes for non–AI-centric digital adopters.
- Strengths and limitations:
- Strengths: mixed methods triangulate mechanisms (qualitative) with causal inference (DiD).
- Limitations: small firm sample (n=8) restricts generalizability and power; short evaluation window (2023–2024); potential selection effects in adoption decisions and heterogeneity across sectors and firm sizes that require broader replication.
Implications for AI Economics
- Causal role of AI architecture: Economic gains from digital transformation for circularity depend on organizational design and decision architectures — modeling AI as an orchestrator clarifies where value is created and captured.
- Investment targeting: Policymakers and firms should prioritize investments in AI decision systems (integration, orchestration, simulation capacity) rather than treating digitization as a checklist of tools; subsidies, grants, or tax incentives could be oriented toward AI orchestration capabilities that yield resource-efficiency externalities.
- Returns to scale and diffusion: The reported effect sizes (18–26% waste reduction, 14–17% reuse gains) suggest substantial private and social returns — cost–benefit analyses and scaled pilot programs in other emerging-market contexts are warranted.
- Labor and organizational implications: AI-centric decision-making reshapes roles (e.g., planners, reverse-logistics coordinators) and requires training, governance, and change management; economic analyses should account for re-skilling costs and productivity trade-offs.
- Data and governance: Effective AI orchestration depends on data quality, interoperability across supply-chain partners, and regulatory frameworks for data sharing and privacy; these public-good elements may justify coordinated policy action.
- Research directions: replicate with larger, diverse samples and longer horizons; disaggregate which AI capabilities (predictive vs. prescriptive vs. simulation) drive which circular outcomes; study distributional impacts, market structure effects, and optimal incentive designs to accelerate adoption in resource-constrained settings.
Assessment
Paper Typequasi_experimental
Evidence Strengthmedium — The DiD design plus qualitative interviews provides plausible causal evidence that AI-centric orchestration drove circular gains, but the quantitative sample is very small (eight firms), the evaluation window is short (2023–2024), and the prompt does not report tests for parallel pre-trends, balance, or robustness checks—raising concerns about statistical power, selection bias, and unobserved confounding.
Methods Rigormedium — Methodologically appropriate mixed-methods approach: interviews strengthen mechanistic claims and DiD is a standard causal technique; however, rigor is limited by the tiny sample size, limited information on covariate adjustment/matching, lack of detail on baseline trends and inference (clustering, standard errors), and potential measurement issues for circular outcomes.
SampleEight manufacturing firms in Mexico and Colombia observed during 2023–2024 for quantitative Difference-in-Differences analysis, plus 17 semi-structured interviews with firm stakeholders (e.g., supply chain managers, engineers, and policymakers) to explore decision architectures and adoption processes; firm sizes, industries, and selection criteria not fully specified in the summary.
Themesinnovation org_design
IdentificationQuasi-experimental Difference-in-Differences comparing circular performance before and after adoption for firms that implemented an AI-centric decision architecture versus a control group of firms that adopted digital but non–AI-centric configurations; supplemented by 17 semi-structured interviews to validate mechanisms and contextualize quantitative results.
GeneralizabilityVery small sample (n=8) limits statistical generalizability, Restricted to two emerging-market countries (Mexico and Colombia); results may not transfer to high-income economies, Manufacturing-sector focus may not apply to services or other industries, Short evaluation window (2023–2024) limits inference about long-run effects and sustainability of gains, Potential selection bias—firms that chose AI-centric architectures may differ systematically from controls, Unclear heterogeneity by firm size, value chain position, or AI maturity reduces external validity
Claims (9)
| Claim | Direction | Outcome | Confidence & Evidence | Details |
|---|---|---|---|---|
| Digital transformation has improved visibility and efficiency in supply chains, yet it has delivered limited progress toward circular economy objectives, particularly in emerging markets. Organizational Efficiency | mixed | visibility and efficiency improvements in supply chains and limited progress on circular economy objectives |
Reading fidelity
high
Study strength
medium
|
n=17
|
| Existing research has largely examined digital technologies in isolation or treated artificial intelligence (AI) as a secondary analytical tool, leaving unclear whether circular performance improvements stem from technology adoption itself or from changes in supply chain decision-making. Other | null_result | clarity on causal source of circular performance improvements (technology adoption vs decision-making) |
Reading fidelity
high
Study strength
medium
|
not reported
|
| This study proposes the AI-Driven Circular Digital Supply Chain (AICD-SC) framework, which conceptualizes AI as a central decision orchestrator integrating predictive, prescriptive, and simulation-based capabilities to coordinate closed-loop supply chain processes. Organizational Efficiency | positive | conceptualization of AI role in coordinating closed-loop supply chain processes |
Reading fidelity
high
Study strength
speculative
|
n=17
|
| The study employs a sequential explanatory mixed-methods design, combining 17 semi-structured interviews with a quasi-experimental Difference-in-Differences analysis of eight manufacturing firms in Mexico and Colombia during 2023–2024. Other | null_result | research design / methods used |
Reading fidelity
high
Study strength
high
|
n=8
|
| Firms adopting AI-centric decision architectures achieved waste reductions of 18–26%. Organizational Efficiency | positive | waste reduction |
Reading fidelity
high
Study strength
medium
|
n=8
18–26%
|
| Firms adopting AI-centric decision architectures achieved improvements in material reuse and recovery of 14–17%. Organizational Efficiency | positive | material reuse and recovery |
Reading fidelity
high
Study strength
medium
|
n=8
14–17%
|
| Firms relying on digitally enabled but non–AI-centric configurations exhibited no statistically significant circular performance gains. Organizational Efficiency | null_result | circular performance gains (e.g., waste reduction, material reuse/recovery) |
Reading fidelity
high
Study strength
medium
|
n=8
|
| These findings indicate that circular outcomes do not emerge from digitalization alone, but from how supply chain decisions are architected and orchestrated through AI. Organizational Efficiency | positive | emergence of circular outcomes (waste reduction, material reuse/recovery) conditional on AI-driven decision architecture |
Reading fidelity
high
Study strength
medium
|
n=8
|
| The study concludes by offering a phased adoption roadmap aligned with Sustainable Development Goal 12, providing actionable implications for managers and policymakers in emerging markets. Governance And Regulation | positive | phased adoption roadmap aligned with SDG12 (policy/managerial guidance) |
Reading fidelity
high
Study strength
speculative
|
n=17
|