3.2IoT, embedded applications & engineering systems

In manufacturing, IoT and engineering systems offer large room to raise production efficiency, quality, and flexibility, and an IT strategist must discern which method to use in which situation. The point here is not to memorize definitions but to judge which production or maintenance method best reconciles efficiency and effectiveness under constraints such as demand variability, supply stability, and product variety. For instance, JIT—which cuts inventory to the extreme—is not universal, and in situations of unstable demand or supply it can instead invite stockout risk. This section covers the representative methods as material for situation-dependent judgment.

3.2.1Production-management and maintenance methods

• MRP (Material Requirements Planning): a plan-driven (push) method that computes required material quantities and order timing from the production plan (master production schedule), bill of materials (BOM), and inventory. It arranges materials in advance based on demand forecasts.
• JIT (Just In Time): a demand-driven (pull) method that makes only what is needed, when needed, in the needed quantity, minimizing inventory. The kanban method, in which the downstream process signals the upstream to withdraw, is the tool that realizes it. Since it holds no inventory, it presupposes stable demand and supply.
• Cell production: a method where a small team or a single worker handles a series of steps together, flexible for high-mix low-volume and spec changes. It suits different situations than conveyor-belt line production, which fits mass production.
• Predictive maintenance: maintenance that acquires equipment condition data via sensors and IoT, detects signs of failure, and acts in advance. Unlike preventive maintenance (periodic time-based replacement) and breakdown (reactive) maintenance (fixing after failure), it services only when needed, avoiding over- and under-maintenance.

3.2.2Smart factory and engineering systems

• The smart factory / Industrie 4.0: a vision of connecting equipment, products, and systems via IoT, optimizing production with collected data, and efficiently producing high variety close to individual orders (mass customization). It is at the core of the German-originated fourth industrial revolution.
• CAD (design) / CAM (manufacturing, NC-data generation) / CAE (analysis, simulation): engineering systems that computer-aid from design through manufacturing and verification. They analyze strength, fluid behavior, etc. before prototyping, cutting design rework and cost.

An IT strategist at a parts maker is reviewing its production and inventory methods. The firm's main line has run with a certain amount of work-in-process (buffer) between each step, but management requests to "fully adopt Toyota-style JIT/kanban to drive inventory near zero and improve cash flow." What the strategist must judge is which—minimal-inventory JIT or an operation holding safety (buffer) stock—optimizes the trade-off between stockout risk and inventory cost, given this product's demand and supply characteristics. JIT's benefit is inventory compression and lead-time reduction, but its precondition for working is relatively stable demand and short-lead-time, high-reliability parts supply. Investigation shows this product's demand fluctuates greatly with seasons and promotions, and some critical parts are procured overseas with long lead times and unstable supply. In this situation, driving inventory uniformly near zero risks production halting the moment demand spikes or a part is delayed, inviting stockout-driven opportunity loss and customer defection—losses far exceeding the inventory saved. The strategist therefore judges not to treat "JIT or safety stock" as either/or, but to use each according to the constraints of demand variability and supply instability. Concretely, the optimum is a hybrid: apply JIT/kanban to squeeze inventory for commodity parts with stable demand and short, reliable supply, while strategically holding statistically computed safety (buffer) stock for critical parts with high demand variability and unstable supply. Alongside, visualizing demand, inventory, and equipment operation in real time via IoT and raising demand-forecast accuracy creates room to lower the safety-stock level itself. The essence here is not the allure of the JIT method but the judgment of optimizing the trade-off against the constraints (demand variability, supply stability).

3.2.3Section summary

• JIT (pull, kanban) minimizes inventory but presupposes stable demand and reliable supply
• Under high demand variability and unstable supply, strategically hold safety (buffer) stock and use it selectively with JIT per the constraints
• Predictive maintenance detects signs from condition data (distinct from time-based preventive and post-failure reactive), and IoT / the smart factory support visualization and optimization