Manufacturing Operations
APICS hierarchy · five levels · fifteen models
Manufacturing operations research asks how a plant, production network, or industrial enterprise can turn raw materials into finished goods with the lowest cost, shortest lead time, and highest quality — across horizons that stretch from an hourly shop-floor dispatch to a decade-long capacity investment. This section presents fifteen canonical OR problems, each as a live interactive solver grounded in a real manufacturing decision, organised along the APICS planning hierarchy — Sales & Operations Planning through Shop Floor Control — grounded in Hax & Meal (1975), Orlicky (1975), Pinedo (2022), and Hopp & Spearman (2011).
Why manufacturing OR matters
Scale of the problem · three anchor statistics
Decision framework
Four lenses on the same fifteen applications
The APICS planning hierarchy stacks manufacturing decisions across five levels, from the yearly Sales & Operations Planning (S&OP) at the top to the hourly Shop Floor Control (SFC) at the bottom. Each level feeds constraints downward and demand forecasts upward. The framework is rooted in the strategic–tactical–operational hierarchy of Anthony (1965), formalised by Hax & Meal (1975), and codified industrially by APICS / ASCM. Click any level to jump to its applications.
The same applications regrouped by OR problem family — the framing used in the scheduling, lot-sizing, and assembly-line-balancing literatures. Anchored in Pinedo (2022) for scheduling, Karimi et al. (2003) for lot sizing, and Scholl (1999) for assembly line balancing.
The same applications regrouped by production-system archetype. Different plants invoke different subsets of this catalog depending on whether they run continuous processes, batches, mass discrete assembly, job-shop custom work, or project-based one-offs. The project column cross-links to construction, which owns resource-constrained project scheduling.
Smart manufacturing (a.k.a. Industry 4.0) wraps classical OR problems in new data, sensing, and control infrastructure — digital twins, real-time telemetry, IoT-driven dispatch, ML-learned dispatch rules, and predictive maintenance. The applications below are those in our catalog that directly plug into smart-manufacturing stacks or whose variants are at the research frontier. See Lu (2017) and Zhong et al. (2017) for surveys.
All fifteen applications
Click any card to open its interactive solver
Current research frontiers
Where manufacturing OR is actively evolving
Deep reinforcement learning for real-time dispatching
Neural-network dispatching rules trained on simulation environments, deployed on cyber-physical shop floors. Early results on the classical Taillard flow-shop and Lawrence job-shop benchmarks are competitive with tabu search at decision times two orders of magnitude lower.
Distributionally robust production planning
Wasserstein and moment-based ambiguity sets around demand, yield, and lead-time distributions — robust CLSP, robust MPS, and data-driven APP models that hedge the tail without over-insuring (Esfahani & Kuhn 2018; Delage & Ye 2010).
Closed-loop & remanufacturing scheduling
Joint production-remanufacturing-disassembly scheduling with return-flow uncertainty, for circular manufacturing and reverse supply chains (Guide, Jayaraman & Srivastava 1999; Souza 2013). Cross-link to logistics for multi-echelon inventory in closed loops.
Additive manufacturing production planning
3D-printing scheduling, part-nesting in build volumes, and distributed-manufacturing routing — a new packing-scheduling hybrid that does not fit classical flow-shop or job-shop templates.
Integrated production-distribution models
Cross-functional MIPs that simultaneously optimise MPS, vehicle routing, and multi-echelon inventory — reducing the two-step bias of plan-then-route approaches (Fahimnia et al. 2013). See logistics for the routing side.
Factory physics at scale
Queueing-network models of entire plants — Jackson networks, mean-value analysis, Hopp & Spearman's VUT equation — re-emerging as the analytical backbone behind digital-twin variability quantification.
Key references
Anchor textbooks & foundational papers · DOIs where available