Finite Capacity Scheduling

The engine builds a schedule by solving a complex constraint satisfaction problem. It simultaneously respects multiple hard limits to prevent overloading.

• Work Center Capacity: Models maximum throughput rates (e.g., units/hour) for each machine.
• Labor & Tooling: Accounts for skilled operator availability and finite tooling sets.
• Material Availability: Synchronizes the schedule with inbound component deliveries.
• Calendar & Shifts: Respects maintenance windows, holidays, and shift patterns.

FCS algorithms can generate a plan in two strategic directions, depending on the business objective.

• Forward Scheduling: Starts from the current date and schedules operations as early as possible to calculate the Earliest Completion Date. Ideal for maximizing utilization.
• Backward Scheduling: Starts from the due date and works backward to calculate the Latest Start Date. Ideal for minimizing work-in-process (WIP) inventory and adhering to lean principles.

When multiple jobs compete for the same resource, the scheduler applies configurable rules to resolve the conflict and sequence the queue.

• Critical Ratio: Prioritizes jobs with the tightest margin between remaining time and remaining work.
• Shortest Processing Time: Minimizes average queue time by running quick jobs first.
• Earliest Due Date: Sequences work purely based on delivery commitments.
• Setup Optimization: Groups similar jobs to minimize changeover times and maximize throughput.

Before releasing a plan to the shop floor, planners can simulate the impact of hypothetical changes without disrupting live operations.

• Scenario Testing: Evaluate the impact of a machine breakdown, a rush order insertion, or a labor shortage.
• Comparative Analysis: Visually compare key performance indicators (KPIs) like on-time delivery and utilization between different scheduling scenarios.
• Bottleneck Prediction: Proactively identify future resource constraints before they cause delays.

The schedule is a living entity. FCS systems ingest real-time data from the shop floor to dynamically adjust the plan when reality deviates from the assumption.

• Event-Driven Triggers: Automatically reschedule when an operation is completed late, a machine goes down, or material fails inspection.
• Incremental vs. Regenerative: Choose between making minor adjustments to the existing plan or completely regenerating the schedule from the current state forward.

The output is typically rendered as an interactive Gantt chart, providing a visual timeline of all operations loaded on each resource.

• Drag-and-Drop Manual Overrides: Planners can manually move operations to accommodate urgent, unplanned work.
• Color-Coded Status: Instantly visualize jobs that are on time, at risk, or already late.
• Dependency Mapping: Clearly see the predecessor-successor links between operations to understand the ripple effect of any delay.

The conceptual opposite of Finite Capacity Scheduling. This method generates a production plan assuming work centers have unlimited capacity, ignoring resource constraints entirely.

• Primary Use: Identifying the theoretical earliest completion date and calculating the gap between demand and available capacity.
• Key Difference: It creates an overloaded plan that must be manually leveled, whereas FCS produces a plan that is executable by design.
• Output: Highlights bottleneck work centers where cumulative load exceeds maximum throughput.

A management philosophy developed by Eliyahu M. Goldratt that directly underpins FCS logic. TOC states that any manageable system is limited by a small number of bottlenecks.

• Drum-Buffer-Rope: The TOC scheduling mechanism where the bottleneck (drum) sets the pace, a time buffer protects it from upstream disruption, and a rope controls material release.
• FCS Connection: Effective FCS systems explicitly model the bottleneck's finite capacity first, then subordinate all non-bottleneck resources to its rhythm.

The class of enterprise software in which Finite Capacity Scheduling engines typically reside. APS systems sit between ERP and shop-floor execution.

• Core Modules: Demand planning, supply network planning, and detailed production scheduling.
• FCS Role: The scheduling module within APS uses genetic algorithms or constraint propagation to solve the complex combinatorial problem of sequencing orders against finite resources.
• Differentiator: Unlike standard ERP, APS can perform simultaneous material and capacity checks in memory.

A superset of FCS that models not only capacity limits but also hard and soft business rules as constraints. The solver finds a feasible sequence that violates no hard constraints.

• Hard Constraints: Machine availability, tooling requirements, material expiration dates.
• Soft Constraints: Preferred sequences, color changeovers, customer priorities. The solver minimizes violations of these.
• Technique: Uses constraint programming (CP) solvers rather than pure mathematical optimization, making it effective for highly complex, rule-heavy environments.

Simple, localized heuristics used to determine which job to process next when a resource becomes free. Often used as a baseline comparison for FCS.

• Common Rules: Earliest Due Date (EDD), Shortest Processing Time (SPT), First-In-First-Out (FIFO).
• Limitation: These rules are myopic; they optimize locally at a single machine without considering the global impact on downstream resources.
• FCS Advantage: A global FCS algorithm considers the entire routing network, avoiding decisions that clear one queue only to create a larger bottleneck elsewhere.

The execution layer that translates an FCS-generated plan into shop-floor reality. PAC manages the release, monitoring, and correction of work orders.

• Input: The dispatch list from the FCS, specifying the exact sequence and start times for each operation.
• Function: Tracks actual progress against the schedule, reports variances (e.g., scrap, downtime), and triggers rescheduling when deviations exceed a defined threshold.
• Feedback Loop: Provides the actual capacity data needed to continuously refine the finite model's accuracy.