{"id":"0597b98e-c48c-4630-be81-111a15ed0a0d","arxiv_id":"2505.00891","paper_version":1,"verdict":"UNVERDICTED","confidence":"HIGH","novelty_score":1.0,"correctness_risk":"low","formal_verification":"none","parameter_count":0,"one_line_summary":"A review of quantum computing for industrial optimization that summarizes existing approaches and introduces a web demonstrator for job-shop scheduling, without presenting new research results.","lead":"This paper surveys how quantum computers might be used in industry, covering quantum chips, annealers, and tensor networks, plus example problems like scheduling and route planning. It also points to a free online job-shop scheduling demo the authors built, but presents no new measurements or algorithms.","discovery_kind":"review","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Section 3's 'viable alternative' conclusion overreaches: the cited proof-of-concept studies are never benchmarked against classical solvers, and Section 1 concedes no true quantum advantage yet.","rationale":"The paper is a clearly labeled survey and demonstrator notice, and it does contain useful caveats: Section 1 explicitly says we are in the NISQ era with no true quantum advantage, and Section 5 discusses the 'quantum winter' and reduced investment. I do not fault the authors for summarizing the literature. However, the Section 3 conclusion is broader than the evidence shown. The cited papers may demonstrate that industrial problems can be mapped to QUBO/annealers or tensor networks and that feasible solutions are obtainable, but the paper never reports a single comparison against Gurobi/CP-SAT or specialized classical heuristics, which are the default industrial tools. Without such baselines, 'viable alternative' is an interpretive gloss, not an empirical finding. The reader's weakest assumption correctly identifies this missing benchmark. My additional point is that tensor-network successes (Section 2.3) are classical algorithms, so aggregating them under 'quantum computing' further inflates the appearance of evidence for NISQ viability. This does not change the appropriate verdict: the manuscript is not an original research contribution and should not receive a research accept/reject verdict; UNVERDICTED remains correct. If the authors add a small benchmark comparing their demonstrator and cited methods to classical solvers, the overclaim could be converted into a defensible claim.","tokens_in":11125,"tokens_out":4422,"duration_ms":45905,"concrete_test":"Use the largest real-world instances reported in [27] (3D bin packing), [32] (JSSP on D-Wave), and [39] (package delivery routing), and re-solve the same problem instances with state-of-the-art classical solvers (e.g., CP-SAT, Gurobi, or the best published specialized heuristics) under identical wall-clock limits and objective definitions. If the classical solvers equal or beat the reported quantum/hybrid solutions on every instance, then Section 3's 'viable alternative' conclusion is not supported; if the quantum/hybrid solutions win on any realistic instance, the concern is resolved.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim in Section 3 ('it is prudent to conclude that quantum computing is emerging as a viable alternative ... at least on a proof-of-concept scale') is load-bearing for the paper's industrial-applications message. The support offered is a list of selected studies, but no comparison against best available classical solvers is provided; the paper's own Section 1 states 'there is not yet a true quantum advantage,' meaning current quantum computers cannot outperform classical supercomputers. Without such a baseline, 'viable alternative' is an assertion, not a demonstrated result. The problem is compounded by Section 2.3, where tensor networks are explicitly described as classical, quantum-inspired methods, yet Section 3 treats them as evidence for quantum computing; success of a classical method does not validate NISQ hardware. A 'proof-of-concept scale' result can show encoding feasibility, but industrial viability requires competitive solution quality and time on realistic instances. Since none of the cited works is shown to have those properties, the conclusion rests on selected success stories, not systematic evidence.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"This manuscript is a survey/position paper on the use of quantum computing in industrial environments. It reviews three computational paradigms: gate-based quantum computers, quantum annealers, and tensor networks, and then discusses selected industrial applications including bin packing, job-shop scheduling, and robot/vehicle routing. The paper also describes an online demonstrator for the job-shop scheduling problem developed by the authors, and it concludes with a discussion of vendor roadmaps (IBM and D-Wave), investment trends, and the 'quantum winter' narrative. The central conclusion, stated in Section 3, is that quantum computing is 'emerging as a viable alternative to tackle complex real-world problems, at least on a proof-of-concept scale.'","tokens_in":11462,"tokens_out":4510,"duration_ms":46246,"significance":"As a survey, the paper is useful as an accessible introduction to three computational paradigms and to a curated set of recent industrial applications. Its main value is organizational: it brings together examples from bin packing, scheduling, and routing, and it makes the authors' JSSP demonstrator visible to a wider audience. The paper does not present original benchmark results, and its evidence is selected rather than systematic. Its central claim, that quantum computing is a 'viable alternative' at proof-of-concept scale, is plausible but not established by the cited studies, which lack comparisons against classical solvers. If properly qualified, the survey could serve as a starting point for practitioners, but in its current form the conclusion overreaches relative to the evidence.","major_comments":[{"comment":"The sentence 'it is prudent to conclude that quantum computing is emerging as a viable alternative to tackle complex real-world problems, at least on a proof-of-concept scale' is the load-bearing claim of the paper, but it is not supported by the evidence presented. None of the cited works is compared against best-in-class classical solvers, and Section 1 itself states that 'there is not yet a true quantum advantage.' The cited studies demonstrate that industrial problems can be encoded and that small instances can be solved, but that is not the same as demonstrating viability, which requires competitive solution quality and runtime on realistic instances. The authors should either qualify the conclusion explicitly, for example by saying that quantum and quantum-inspired methods are 'objects of active research' rather than viable alternatives, or add a critical comparison of quantum/quantum-inspired results with classical baselines on common benchmark instances.","section":"Section 3, last paragraph"},{"comment":"The paper introduces tensor networks as 'quantum-inspired' classical methods that run on classical devices, yet in Section 3 it includes tensor-network works [24]-[26] among the evidence for the potential of quantum computing in industry. This conflates quantum-inspired classical computation with quantum hardware: a classical method's success cannot validate the prospects of NISQ hardware. The ambiguity affects the central claim, because the 'viable alternative' conclusion could be read as referring either to genuine quantum computers or to a mix that is largely classical. The paper should consistently distinguish these two categories when drawing conclusions, or explicitly extend the claim to 'quantum and quantum-inspired methods.'","section":"Section 2.3 and Section 3"},{"comment":"The text states that 'projections for the maximum number of next-generation qubits (through modular coupling) have been moderated to 1,092 qubits by 2028,' and then in the same paragraph introduces 'Starling and Flamingo' chips with '200 and 2,000 qubits, respectively.' As written, a 2,000-qubit chip appears to contradict the 1,092-qubit projection. The authors should clarify whether the 1,092 figure applies only to a particular class of modular systems, and they should specify the timeline and system configurations to which each number refers.","section":"Section 5, IBM roadmap"}],"minor_comments":[{"comment":"The figure numbering is inconsistent: the demonstrator page and the IBM roadmap are both numbered Figure 7, and the D-Wave roadmap is numbered Figure 8. The figures should be renumbered sequentially.","section":"Section 4 and Section 5"},{"comment":"The footnote repeatedly uses 'cubits' instead of 'qubits'; this typo should be corrected throughout the footnote.","section":"Footnote 2"},{"comment":"The descriptions of references [29] and [30] appear to be swapped relative to their titles: [30] is 'Aircraft Loading Optimization,' while the text attributes cargo-plane loading to [29] and spent nuclear fuel storage to [30]. Please verify and correct this attribution.","section":"Section 3, bin packing paragraph"},{"comment":"References [49]-[51] are listed in the bibliography but are not cited anywhere in the text; they should either be cited in the relevant discussion of European initiatives or removed.","section":"References"},{"comment":"The only tensor-network industrial examples cited in Section 3 are the authors' own papers [24]-[26]. Including independent case studies would strengthen the survey's credibility and reduce the appearance of selection bias.","section":"Section 3, tensor-network applications"}],"recommendation":"major_revision","confidential_remarks":"The paper is a survey rather than an original technical contribution, and its main claim is not sufficiently supported by the evidence as currently presented. The lack of classical baselines is the key issue; the authors should either temper the conclusion or add a comparative analysis. The repeated internal inconsistencies (duplicate figure numbers, swapped reference attributions, the contradictory qubit projections) suggest a rushed revision, but all of these are fixable within the manuscript's scope."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Colleague,\n\nThis is a review article, not a research paper. It's an English reprint of a work published elsewhere, summarizing three quantum-computing paradigms and their industrial applications. If you're looking for new results, there are none: no equations, datasets, or benchmarks. What it does offer is a clear, accurate introduction to the area, with a genuinely useful web demonstrator for the Job Shop Scheduling Problem that runs on simulators and real D-Wave hardware. That demo is a nice outreach artifact.\n\nThe strong point is that Section 1 is honest: \"there is not yet a true quantum advantage.\" The survey also covers the main paradigms and cites primary literature. The authors' self-citations (tensor-network papers, routing) are descriptive and not load-bearing. Fine.\n\nThe soft spot is the conclusion in Section 3: \"it is prudent to conclude that quantum computing is emerging as a viable alternative to tackle complex real-world problems, at least on a proof-of-concept scale.\" That is an overreach. The cited studies are selected success stories; no comparison against best classical solvers is presented. The stress-test note is correct: a proof-of-concept can show encoding feasibility, but not industrial viability. Even the qualifier doesn't rescue it, because the reader has no idea whether these solutions are competitive. Also, the section treats tensor networks as evidence for quantum computing, when Section 2.3 correctly says they are classical, quantum-inspired methods.\n\nA second issue is internal inconsistency in Section 5. The text says next-generation qubit projections have been \"moderated to 1,092 qubits by 2028,\" then immediately says Flamingo will have 2,000 qubits. That needs fixing. Also there are duplicate figure numbers. Minor editorial stuff, but the qubit contradiction is substantive.\n\nOverall, the paper is fine as a survey for readers new to the area. The central claim should be softened, and the roadmap paragraph cleaned up. It does not deserve peer review as a research contribution; if submitted to a review venue, it could be acceptable after revision. Would I cite it? Probably not. For a reading group, it's a passable overview but not a must-read.","headline":"A readable survey with a modest outreach tool; the 'viable alternative' claim rests on anecdotes, and the qubit roadmap passage contradicts itself.","tokens_in":11860,"tokens_out":2156,"would_cite":false,"duration_ms":20819,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":false},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"Quantum computing is a viable industrial tool, review concludes","keywords":["quantum computing","industrial optimization","quantum annealing","gate-based quantum computers","tensor networks","job shop scheduling","bin packing","vehicle routing"],"falsifier":"Run the hybrid quantum, annealing, and tensor-network methods from the cited studies on standard public benchmark instances of bin packing, job-shop scheduling, or vehicle routing, alongside best-in-class classical solvers; if the classical solvers match or beat them on solution quality and time across the board, the 'viable at proof-of-concept scale' claim would lose its support.","tokens_in":10973,"feed_emoji":"⚛️","tokens_out":4535,"duration_ms":42210,"temperature":0.7,"pith_summary":"This review argues that quantum computing is becoming a viable alternative for solving real industrial problems, at least at proof-of-concept scale. Drawing on published studies of bin packing, job-shop scheduling, and route planning, it walks through three computational paradigms—gate-based quantum computers, quantum annealers, and tensor networks—and shows how hybrid classical-quantum methods have produced working solutions on current NISQ hardware. The authors also present a free online demonstrator for the job-shop scheduling problem that lets non-experts run these methods. The paper is careful to note that no true quantum advantage has been demonstrated yet, but it sees the breadth of proof-of-concept results as evidence that the field is moving toward industrial utility.","feed_headline":"Quantum computing is a viable industrial tool, review finds","feed_subtitle":"Proof-of-concept studies of bin packing, scheduling, and routing point to real value for industry despite no true quantum advantage yet.","key_machinery":"The argument is carried by the three computational paradigms it surveys—gate-based quantum computers, quantum annealers, and tensor networks—plus the hybrid classical-quantum design shared by most of the cited applications. In the hybrid approach, a classical routine decomposes the industrial problem (clustering customers, assigning jobs to machines) and the quantum device solves the resulting subproblems, whether through QAOA/VQE circuits, a quantum annealer evolving an Ising Hamiltonian, or a tensor network that represents all candidate solutions and discards infeasible ones. The paper also introduces its own online JSSP demonstrator, which exposes these methods to non-specialist users, as a practical vehicle for the same idea.","core_discovery":"The paper's central claim is that quantum computing is emerging as a viable alternative to tackle complex real-world problems 'at least on a proof-of-concept scale.' It grounds this claim in a selection of recent studies: hybrid algorithms on D-Wave annealers solve realistic three-dimensional bin packing and delivery-route problems; QAOA and annealing implementations address job-shop scheduling; and tensor-network methods compress and generate data tensors for combinatorial optimization, including a traveling-salesman network. The authors read this accumulated activity as indicating that NISQ-era devices, while not yet beating classical supercomputers, can already provide value in specific industrial tasks. They further argue that quantum computers are special-purpose devices that will complement, not replace, classical computers.","pith_inferences":["One implication the authors do not spell out: even without a quantum speedup, hybrid methods may win in practice if the annealer or circuit quickly produces high-quality solutions that a classical local search then refines; this is testable today with open benchmark instances.","The survey's case-study selection suggests a concrete experiment: take standard bin-packing or JSSP instances from the literature, run the paper's cited hybrid methods against the best classical solvers, and record time-to-solution and solution quality; the result would quantify the gap the paper leaves open.","If the online demonstrator is extended with such classical baselines, it could serve as a neutral testbed for the 'viable at proof-of-concept scale' claim, letting readers see directly where quantum and quantum-inspired methods win or lose.","The comparison between China's large public investment and North America's private-investment drop hints that the next industrial quantum advances may be geographically uneven, a trajectory the paper mentions but does not analyze."],"forward_implications":["If the proof-of-concept claim holds, industrial early adopters can already experiment with hybrid quantum methods on real logistics, manufacturing, and scheduling instances without waiting for fault-tolerant hardware.","The pattern of results suggests that quantum annealers, despite lacking universality, are the most immediately usable hardware class for combinatorial optimization in industry.","Tensor-network methods, being classical, offer a low-risk way to gain experience with quantum-inspired formulations; their successes (e.g., 93% compression of large language models) strengthen the case that quantum-derived approaches can pay off before quantum hardware matures.","Continued investment in NISQ-era tools is justified even while the field faces a 'quantum winter' of reduced private funding, since the proof-of-concept work documented here does not depend on fault tolerance.","IBM's strategy shift from qubit count to qubit quality and modular coupling implies, if the paper's reading is right, that progress in quantum industrial applications will come from error mitigation and hybrid workflows rather than raw qubit counts."],"supporting_citations":[{"why":"Defines the NISQ era and frames the acknowledged limitation that no true quantum advantage exists yet.","marker":"[7]"},{"why":"Supplies the tensor-network exemplar for combinatorial optimization via a traveling salesman problem network.","marker":"[26]"},{"why":"Flagship bin-packing case study using a hybrid D-Wave annealer with real-world constraints.","marker":"[27]"},{"why":"Gate-based QAOA implementation for job shop scheduling, covering the circuit paradigm.","marker":"[31]"},{"why":"Quantum annealing implementation of job shop scheduling executed on D-Wave devices.","marker":"[32]"},{"why":"Multi-robot route planning hybrid that uses D-Wave with classical accept/reject/correct analysis.","marker":"[35]"},{"why":"Capacitated vehicle routing solved by classical clustering plus quantum-annealed traveling salesman subproblems.","marker":"[37]"},{"why":"IBM's published qubit-count roadmap that anchors the paper's discussion of hardware scaling and strategic shifts.","marker":"[41]"},{"why":"D-Wave Advantage2 prototype specifications used to project near-term annealer capabilities.","marker":"[43]"}],"fun_headline_variants":["Quantum computing: proof-of-concept wins for industrial scheduling","Industrial quantum: bin packing and routing show NISQ promise","Quantum for industry: special-purpose tool, not yet advantage","No advantage yet, but quantum tackles real industrial problems","Quantum in industry: proof-of-concept value, no quantum advantage"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The conclusion rests on the assumption that the selected proof-of-concept studies are representative of what quantum and quantum-inspired methods can do in real industrial settings, even though the paper itself states that no true quantum advantage has yet been shown and presents no comparison against the best classical solvers.","fun_headline_variants_meta":{"raw":{"variants":["Quantum computing: proof-of-concept wins for industrial scheduling","Industrial quantum: bin packing and routing show NISQ promise","Quantum for industry: special-purpose tool, not yet advantage","No advantage yet, but quantum tackles real industrial problems","Quantum in industry: proof-of-concept value, no quantum advantage"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000242,"raw_usage":{"total_tokens":1448,"prompt_tokens":794,"completion_tokens":654,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":410,"completion_tokens_details":{"reasoning_tokens":573}},"tokens_in":410,"tokens_out":654,"duration_ms":6632,"temperature":1.0,"reasoning_tokens":573,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-16T04:31:50.046333+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Run the hybrid quantum, annealing, and tensor-network methods from the cited studies on standard public benchmark instances of bin packing, job-shop scheduling, or vehicle routing, alongside best-in-class classical solvers; if the classical solvers match or beat them on solution quality and time across the board, the 'viable at proof-of-concept scale' claim would lose its support.","supporting_citations":[{"cited_title":"V., Osaba, E., Villar-Rodriguez, E., Oregi, I., & Ban, Y","cited_arxiv_id":null,"evidence_quote":"Flagship bin-packing case study using a hybrid D-Wave annealer with real-world constraints."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Gate-based QAOA implementation for job shop scheduling, covering the circuit paradigm."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Quantum annealing implementation of job shop scheduling executed on D-Wave devices."},{"cited_title":"(2019, January)","cited_arxiv_id":null,"evidence_quote":"Multi-robot route planning hybrid that uses D-Wave with classical accept/reject/correct analysis."},{"cited_title":"& Linnhoff-Popien, C","cited_arxiv_id":null,"evidence_quote":"Capacitated vehicle routing solved by classical clustering plus quantum-annealed traveling salesman subproblems."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"D-Wave Advantage2 prototype specifications used to project near-term annealer capabilities."}],"review_version":1}