Sense
Observe changing demand, supply conditions, context, and externalities.
Duke University · The Fuqua School of Business
R. David Thomas Professor of Business Administration
Professor of Operations Management
The science of matching resources, capabilities, and evidence to changing demand.
How should organizations sense changes in demand, supply, and context—and respond with the right resources, at the right time, under uncertainty?
My research develops the states, structural principles, and governance mechanisms that connect observation to action. The resource may be inventory, capacity, flexibility, information, human expertise, or decision-ready data; the objective is to help supply systems prepare, respond, learn, and adapt as uncertainty unfolds.
Research lens
The five areas below organize the published and ongoing research by intellectual contribution. Select a topic to see its research question, central insights, complete journal-publication list, and related working papers; the CV, Google Scholar, and SSRN retain the chronological record.
Observe changing demand, supply conditions, context, and externalities.
Identify the decision state, service unit, and information structure.
Preserve inventory, capacity, flexibility, relationships, and other options.
Allocate, replenish, expedite, configure, price, contract, and coordinate.
Use outcomes to update policies, data, contracts, and network capabilities.
Outcomes become the next signals, closing the loop.
Sense the state · Position inventory across stages
How should inventory be controlled at a single stocking point and positioned across a multi-stage network when demand, supply conditions, and lead times evolve over time?
At both levels, useful policy structure survives uncertainty when the state representation matches the operating timeline: inventory position and order coverage at a single stage, and echelon states and decompositions across a network.
Demand regimes, supply conditions, lead-time distributions, outstanding orders, obsolescence, capacity, and each location’s position in the network.
Base-stock, (r,q), and (S,T) policies; order-coverage logic; echelon controls; stochastic comparisons; and scalable bounds and approximations.
Subcategory 1
How should one stocking point set replenishment, capacity, service, and pricing decisions under changing demand, uncertain supply, stochastic lead times, and obsolescence?
A compact state—inventory position together with the relevant demand or supply state—supports interpretable base-stock, (r,q), (S,T), queueing, and pricing policies.
Subcategory 2
How should inventory be positioned and controlled across serial and distribution networks when downstream service depends on decisions made at several stages?
Echelon inventories and carefully chosen decompositions preserve optimal structure and convert high-dimensional network problems into scalable bounds, heuristics, and local calculations.
Single-stage inventory models
Manufacturing & Service Operations Management 28(1), 255-270.
Manufacturing & Service Operations Management 22(1), 36-46.
Operations Research 65(5), 1414-1428.
Production and Operations Management 25(9), 1513-1527.
European Journal of Operational Research 229, 95-105.
Naval Research Logistics 59(8), 601-612.
Manufacturing & Service Operations Management 14(1), 42-49.
Advances in Applied Probability 43(1), 264-275.
Operations Research 58(1), 68-80.
European Journal of Operational Research 111, 28–49.
Management Science 42, 1409-1419.
Management Science 42, 1352-1363.
Operations Research 44, 215-222.
Operations Research Letters 15, 85-93.
Management Science 40, 603-613.
Operations Research 41, 351-370.
Scientia Sinica (Series A) XXXI, 1281-1291.
Science Bulletin 33, 448-454.
Science Bulletin (Chinese Edition) 32, 1201-1205.
Multiechelon inventory models
Manufacturing & Service Operations Management 24(4), 2310-2327.
Operations Research 65(5), 1231-1249.
Operations Research 55(5), 843-853.
Manufacturing & Service Operations Management 8(4), 394-406.
Operations Research 53(2), 350-362.
IMA Journal of Management Mathematics 14, 321-336.
Manufacturing & Service Operations Management 5, 372-374.
Management Science 49(5): 618-638.
Operations Research 49(2), 226-234.
Naval Research Logistics 43, 381-396.
Naval Research Logistics 39, 715-728.
No research items match this filter.
Sense the customer order · Respond through shared components
When final products share components and are assembled after demand arrives, what is the right unit of service and how should components be stocked and allocated across customer orders?
The customer order—not the individual component—is the natural service unit. Shared components create dependence, but bills of materials and order structure also make exact decompositions, performance bounds, and scalable policies possible.
Bills of materials, customer-order composition, component commonality, order waiting, stochastic component lead times, returns, and advance demand information.
Order-based service measures, component base-stock decisions, no-holdback allocation, postponement, commonality design, and asymptotically optimal policies.
Working papers
Published and forthcoming
Manufacturing & Service Operations Management 26(6), 2194-2211.
Naval Research Logistics 62(8), 617-645.
Operations Research 58(3), 691-705.
Manufacturing & Service Operations Management 11(3), 493-508.
Manufacturing & Service Operations Management 11(1), 144-159.
IIE Transactions 37(8), 763-774.
Operations Research 53(1), 151-169.
Manufacturing & Service Operations Management 5, 230-251.
Operations Research 51, 292-308.
Operations Research 50, 889-903.
Management Science 48, 499-516.
Manufacturing & Service Operations Management 2, 287-296.
Management Science 46, 739-743.
Operations Research 47, 131-149.
Operations Research 46, 831-845.
No research items match this filter.
Sense disruptions · Preserve and deploy options
What options should a supply system preserve before uncertainty is resolved, and how should it deploy sourcing, expediting, capacity, pooling, and inventory after disruptions?
Resilience is a portfolio of timed operational options, not simply more stock. Pipeline visibility can turn stochastic supply into decision-relevant state information, while access and retrieval rules can create flexibility without duplicating every physical resource.
Disruptions, demand surges, real-time order locations and status, capacity congestion, supply reliability, and service urgency.
Dual and multisourcing, expediting, virtual stockpile pooling, prepositioning and local purchasing, returns, reactive capacity, and operations reversal.
Working papers
Working paper.
Working paper.
Working paper.
Working paper.
Working paper.
Working paper.
Working paper.
Working paper.
Published and forthcoming
Operations Research. Article in advance.
Operations Research 73(6):2867-2885.
Fundamental Research 5, 450-463.
Operations Research 70(4), 2421-2438.
Production and Operations Management 31(5), 2015-2037.
Manufacturing & Service Operations Management 24(1), 315-332.
Health Care Management Science 24(3), 455-459.
Operations Research 65(2), 379-395.
European Journal of Operational Research 259(1), 100-112.
Manufacturing & Service Operations Management 18(4), 509-524.
Production and Operations Management 25(10), 1745-1762.
Journal of Applied Business and Economics 18(1).
Manufacturing & Service Operations Management 15 (3), 444-457.
Management Science 55(3), 362-372.
No research items match this filter.
Sense incentives and externalities · Respond through governance
How should decentralized supply networks allocate information, authority, risk, and rewards when operational objectives interact with social and environmental responsibility?
Coordination depends on payment timing, decision rights, delegation, auditing, and participation—not price alone. More direct control or stronger external pressure is not always better when local information, hidden actions, and multi-tier incentives matter.
Private demand and cost information, multi-tier responsibility risk, local knowledge, compliance conditions, financial exposure, and channel incentives.
Contracts, dynamic mechanisms, payment timing, guided delegation, audits, ownership and channel structures, and incentive-compatible partnerships.
Working papers
Working paper.
Working paper.
Working paper.
Working paper.
Working paper.
Published and forthcoming
Management Science. Forthcoming.
Also listed under Digital, data & AI.
Manufacturing & Service Operations Management 25(6), 2314–2332.
Manufacturing & Service Operations Management 24(6), 2843-2862.
Manufacturing & Service Operations Management 23 (2), 294-310.
Manufacturing & Service Operations Management 22(2), 346-363.
Operations Research 67(4), 984-1001.
Manufacturing & Service Operations Management 21(2), 435-451.
Manufacturing & Service Operations Management 21(2), 452-467.
Production and Operations Management 26(7), 1268-1283.
Production and Operations Management 26(2), 305-319.
European Journal of Operational Research 258(2), 525-536.
International Journal of Production Economics 128(1), 175-187.
Naval Research Logistics 56(8), 745-765.
Management Science 55(4), 685-695.
European Journal of Operational Research 187(3), 671-690.
IIE Transactions 39, 111-124.
Manufacturing & Service Operations Management 6(1), 53-72.
No research items match this filter.
Sense through technology · Respond through learning and action
How do digital production, sensors, indices, transparency, operational data, and AI change what firms can observe and the actions they can take?
Technology creates value only when it changes a feasible response and when incentives support that response. Learning and control therefore need to be designed jointly rather than treating prediction as separate from operations.
Sensor signals, freshness, transaction and operational data, market conditions, inventory shrinkage, equipment health, changing demand, and data-derived yield indices.
Stock-or-print decisions, predictive printing, distributed production and licensing, data-driven pricing and replenishment, smart contracts, index-based risk protection, and AI-enabled quick response.
Working papers
Working paper.
Working paper.
Working paper.
Working paper.
Published and forthcoming
Management Business Review. Forthcoming.
Management Science. Forthcoming.
Also listed under Coordination & responsibility.
Manufacturing & Service Operations Management. Article in advance.
Management Science. Forthcoming.
Management Science, 71(8), 6666-6682.
Management Science 71(3), 1925-1943.
INFORMS Transactions on Education 25(1), 85-88.
INFORMS Transactions on Education 25(1), 81-84.
Manufacturing & Service Operations Management 25(4), 1338–1356.
Manufacturing & Service Operations Management 24(5), 2685-2702.
Production and Operations Management 31(6), 2477-2491.
Management Science 68(3), 1938-1958.
Supply Chain Management Review, November: 28-33. Academic paper (e-companion): 3D printing spare parts: A scalable total-cost framework for sourcing decisions.
Operations Research 69(2), 525–544.
Management Science 66(9), 3860-3878.
Operations Research 65(6), 1574-1588.
Operations Research 58(4), 1034-1038.
Operations Research 54(6), 1079-1097.
Operations Research 53(6), 1024-1026.
No research items match this filter.
Selected research questions and findings
Rather than repeating the chronological publication record, these summaries foreground the operating question and the insight each paper contributes.
Books
These books range from a research handbook on inventory theory, to an edited volume on supply-chain structures, to a Chinese-language introduction that uses everyday life to explain the art of matching products, services, and processes.
A comprehensive reference on quantitative inventory models. The chapters review foundational theories and methods, interfaces between inventory and other business decisions, and industry-specific challenges in healthcare, spare-parts logistics, retailing, and online retailing.
Publisher information ↗The Art of Matching: Joy of Living and Operations Innovations
Written as conversations between two protagonists, the book uses examples from clothing, food, housing, and transportation to explain how organizations match products and services with processes—and how operational innovation improves everyday life.
Book information ↗An edited collection connecting supply-chain architecture with coordination, information sharing, manufacturing flexibility, assemble-to-order planning, inventory allocation, and industry cases. It brings together academic and practitioner perspectives on how supply chains should be designed and managed.
Springer book page ↗Research in the media
Selected features, interviews, and practitioner pieces explain how the research applies to adaptive supply chains, sustainability, food freshness, and digital spare-parts operations.
Follow research updates on LinkedIn ↗How AI is moving supply chains from fixed workflows toward systems that sense, learn, adapt, and act—while raising new questions about human judgment, accountability, and governance.
Read feature ↗Why responsibility failures in distant tiers require incentive design, auditing, and governance—not simply stronger statements of intent.
Read feature ↗How freshness transparency can improve ordering and reduce food waste—and why a smart contract may be needed to share the gains.
Read feature ↗A practitioner-oriented framework for comparing additive and traditional manufacturing across the full part lifecycle and large SKU portfolios.
Read article ↗A research conversation on when original equipment manufacturers should stock parts, print them on demand, or combine the two modes.
View talk listing ↗Industry coverage of the stock-or-print model and its practical message: a lightly used printing option can still eliminate substantial inventory.
Read coverage ↗Teaching & mentoring
DissertationSimple Solutions to Supply Chain Inventory Management
PlacementDuke University, Fuqua School of Business
DissertationEssays on Multi-Channel Supply Chains
PlacementPeking University, Guanghua School of Management
DissertationIntegrated Marketing and Operations Strategies: Multiproduct Demand Shaping and Inventory Planning
PlacementIBM Watson Research Center
DissertationCoordination Mechanism Design for Sustainable Global Supply Networks
PlacementNanyang Technological University, Nanyang Business School
DissertationInventory Management and Supply Chain Finance: Theory and Empirics
PlacementUniversity of Wisconsin-Madison, Wisconsin School of Business
DissertationCoopetition in a Supply Chain
PlacementShangdong University, School of Business
DissertationResponsible Sourcing and Supply Chain Risk Management
PlacementGoogle Inc., Operations Decision Support
DissertationEssays in Empirical Operations Management: Bayesian Learning of Service Quality and Structural Estimation of Complementary Product Pricing and Inventory Management
PlacementFacebook, Inc.
DissertationGlobal Supply Chain Management with Advanced Information and Production Technologies
PlacementPennsylvania State University, The Smeal College of Business
DissertationEffective Heuristics for Dynamic Pricing and Scheduling Problems with High Dimensionality
PlacementT-Mobile
DissertationData-driven Decision Making with Dynamic Learning under Uncertainty: Theory and Applications
PlacementJohns Hopkins Carey Business School
DissertationDesign and Performance Prediction for Supply Chain Systems with Graphical Structures
PlacementAmerican Airlines, Operations Research & Advanced Analytics
DissertationIndex-Based Yield Protection Policies: Insights and Data-Driven Policies
PlacementGraham Capital Management
DissertationTechnology and Operations Management for Sustainability and Social Responsibility
PlacementUniversity of Alberta, Alberta School of Business
Scholarly leadership
My service has focused on strengthening the intellectual standards, research communities, and cross-border conversations that sustain operations management as a field.
Jing-Sheng Jeannette Song is the R. David Thomas Professor of Business Administration and Professor of Operations Management at Duke University’s Fuqua School of Business. Her research spans single-stage and multiechelon inventory theory, assemble-to-order systems, flexibility and resilience, coordination and responsible operations, digitization, data-driven decision making, and AI-enabled supply systems.
Jing-Sheng Jeannette Song is the R. David Thomas Professor of Business Administration and Professor of Operations Management at Duke University’s Fuqua School of Business. Her research spans supply-chain inventory optimization, assemble-to-order and multiechelon systems, flexibility and resilience, coordination and responsible operations, digitization, data-driven decision making, and AI-enabled supply systems.
She has published more than ninety refereed journal articles; edited the Research Handbook on Inventory Management; co-edited Supply Chain Structures: Coordination, Information and Optimization; and co-authored the Chinese-language book The Art of Matching: Joy of Living and Operations Innovations.
PhD, Management Science, Columbia University
MS, Operations Research, Chinese Academy of Sciences
BS, Mathematics, Beijing Normal University
jssong@duke.edu
The Fuqua School of Business
Duke University
100 Fuqua Drive, Durham, NC 27708