EDGEBIC: The Complete Guide to Next-Generation Finite Capacity Scheduling

EDGEBIC by User Solutions is a finite capacity planning and scheduling platform that turns your open manufacturing orders into a feasible, machine-by-machine, shift-by-shift production schedule. It is the next-generation successor to RMDB and EDGEBI, built by User Solutions, Inc., the company that has been solving production scheduling problems for manufacturers since 1991. Where an ERP tells you what you owe and what you sold, EDGEBIC answers the harder daily question: what should run on which machine, when, and in what order, given the capacity you actually have.

This guide is the complete map. It covers where EDGEBIC sits in the planning stack, what finite capacity means in practice, the master data you set up once, the scheduling engine and every major capability inside it, shop-floor execution, the optimizer, quoting, and how to get from a downloaded installer to your first committed schedule. For a deeper look at any single topic, follow the links into the rest of the EDGEBIC library.

Where EDGEBIC Sits in the Planning Stack

Manufacturing planning has a strict lineage, and the terms are not interchangeable marketing words. A Master Production Schedule (MPS) states what finished goods to build and by when. MRP explodes bills of material, nets against inventory, and offsets lead times, but it plans on infinite capacity: it asks what material is needed, never whether the machine can actually do the work that week. An Advanced Planning and Scheduling (APS) system closes that gap by scheduling material and machine capacity together, at finite capacity, into an executable plan with real start and end times.

EDGEBIC is an APS engine with the surrounding layers a mid-size manufacturer needs: master data, quoting with what-if simulation, shop-floor actuals capture, reschedule logic, reports, and dashboards. It also reaches up the stack with inventory, forecasting, and Master Production Schedule capability for make-to-stock products, so a stocked bracket and a custom one-off order can be planned in the same run. It is not an ERP: it does not run your general ledger or your purchasing. It connects to the system that does through flexible Excel, CSV, and database import-export masks.

Here is the difference finite capacity makes, using a one-booth paint shop as the example. Three jobs land on the same Monday: 6 hours of white, 4 hours of red, 8 hours of black. That is 18 hours of demand against an 8-hour day.

The infinite-capacity answer is a to-do list. The finite-capacity answer is a schedule. If that distinction is new territory, start with finite vs infinite capacity scheduling and come back.

The Company Behind It

User Solutions, Inc. was founded in 1991 and has spent 35+ years in production scheduling. Its customer list includes the US Navy, GE, BAE Systems, and Cummins. The results from that lineage are concrete: GE Railcar took on-time delivery from 30% to 90%, the USS Nimitz coordinated 26,000+ tasks on User Solutions scheduling technology, and Cummins deployed across 33 locations. When Plastilite needed its scheduler integrated with Fourth Shift ERP, the vendor-recommended integration was live in 5 days.

EDGEBIC carries that heritage onto a modern platform: a .NET 8 Windows application with a graphical routing designer, a rich Gantt, a shop-floor kiosk, and a database that can be a single local file or a shared SQL Server. If you run RMDB or EDGEBI today, the upgrade path keeps your data, your routings, and your scheduling concepts.

The Master Data: Describe Your Plant Once

Everything EDGEBIC does rests on four kinds of master data. You set them up once and maintain them occasionally. The First Time Setup Wizard walks you through them in dependency order, and every one of them can also be loaded from Excel or CSV through an import mask.

Shifts and calendars. A shift is a named recurring time block: Day Shift, Monday to Friday, 08:00 to 16:00 with a one-hour break, netting 7 working hours per day. Layer on plant holidays, work-center-specific holidays (a maintenance window on one machine), and downtime events. The scheduler never books an hour the calendar says does not exist.

Work centers. A work center is a machine, cell, or resource with real capacity: a number of instances (two identical mills inside one work center), a utilization percentage, its own shift assignments, and flags that change scheduling behavior. Mark a furnace as one-job-per-instance-per-day, mark your constraint as the bottleneck for Theory of Constraints scheduling, or mark a paint line as a continuous process.

Products. Each product carries its planning attributes: make-to-order or make-to-stock, lead time, yield, safety stock, and reorder rules for stocked items.

Bills of Routing (BOR). The routing is the sequence of steps that makes a product: which work center, hours per unit, setup time, queue time, and what comes next. You can build a routing in a grid or draw it as a flow chart in the graphical designer, with sub-assemblies, branches, and parallel legs. Steps carry the advanced knobs too: alternative work centers, transfer batches for lot streaming, transit days for outside processing, a required operator skill. The how-to for building one visually lives in the routing flow-chart guide.

The Scheduling Engine

The engine is the heart of EDGEBIC. You create manufacturing orders (a product, a quantity, a due date, a priority), click Schedule, and the engine places every routing step onto real machine time. A full tour lives in Inside a Modern Finite Capacity Scheduling Engine; here is the shape of it.

Scheduling modes

Not every run should touch every job. EDGEBIC gives you five modes:

How a job gets placed

For each order, the engine builds a dependency graph from the routing, sorts the steps so no step is ever placed before its predecessors, then schedules each step forward: the step's start is the latest end among everything it depends on, and its hours are allocated shift by shift, day by day, until the work fits. A 12-hour operation on a machine with one 8-hour shift takes 8 hours Monday and 4 hours Tuesday, ending Tuesday at noon. Capacity per shift is computed as shift hours, times number of instances, times the work center's utilization percentage, minus downtime and partial holidays.

Jobs are processed in priority order (lower number first, then start date, then due date), so your most important order gets first claim on contested capacity.

The advanced capabilities

Each of these is a documented, working part of the engine, and each has its own deep-dive post:

• Theory of Constraints anchor scheduling. Pin your bottleneck operation to a target date and the engine schedules upstream steps backward to feed it and downstream steps forward from it, with optional constraint, feeding, and shipping buffers. See the bottleneck anchor walkthrough.
• Backward (just-in-time) scheduling. Mark an order to schedule backward and the engine right-aligns it so the last step ends at the due date, starting everything as late as possible. If it does not fit, it falls back to forward scheduling and tells you.
• Sequence-dependent setup matrix. A paint booth needs 5 minutes to go white to cream and 240 minutes to go black to white. The setup matrix stores changeover time per from-product, to-product pair (grouped into families so 100 products need 64 cells, not 10,000), and the engine resolves the true changeover for every job based on what ran before it. The paint shop example shows setup falling 73% from sequencing alone.
• Parallel and alternative work centers. A step can fall back to a different machine when the primary is busy, split across independent machines running simultaneously, or mirror across synchronized machines that must start and finish together, like the three-spindle drilling example.
• Work center groups. Route a step to "any member of the MILLING pool" instead of one machine, with per-member speed factors, and let the engine shop the pool on every run. Add a fourth mill to the group and every routing that uses it benefits immediately. See what a work center group is.
• Operators and skills. Machines are not the only constraint. Define operators, certifications with expiry dates, weekly shift rosters, and time off, and require a skill on a routing step. If your only certified welder is on vacation Thursday, the Inconel job slides, and the schedule tells you before the floor finds out.
• Lot streaming. Let a downstream step start before the upstream step finishes the whole lot, either by piece count (start drilling once 100 pieces are cut) or by a start-to-start lag in hours. The lot streaming walkthrough shows the makespan compression; the concept post on transfer batches explains the theory.
• Queue time and transit days. Model the buffer between operations (counted only during working shifts) and the days a part spends at an outside heat-treater or another plant.

Execution: Actuals, the Kiosk, and Rescheduling

A schedule is only useful if it survives contact with the shop floor. EDGEBIC closes the loop three ways.

The shop-floor kiosk. Operators punch in and out of operations on a simplified terminal, launched from the main toolbar as a separate window. Punches become actual start dates, actual end dates, and logged hours on the schedule, without anyone touching the planning screens.

Actuals in the Job View. Planners can log actual dates and hours directly, mark steps complete, and watch planned-versus-actual on the Gantt.

Reschedule with actuals preservation. This is the invariant the whole system is built around: a completed operation is never moved by a reschedule. When you reschedule a job that is partly done, the engine locks the completed steps as history, takes the partially completed step's remaining hours only (an operation planned at 19.2 hours with 4 logged reschedules just the remaining 15.2), and cascades the untouched steps forward from where the floor actually is. Your history stays true; your future plan stays honest. The partial completion walkthrough traces it step by step.

The Optimizer: Proposals, Never Surprises

The scheduling engine is greedy: it places each job at its earliest feasible slot, one job at a time. That is fast and respects every constraint, but the dispatch order it picks is not always the best one. The optimizer attacks exactly that, and it does so as a proposal engine, never behind your back.

Two selectable layers ship:

• Multi-run search re-runs the real engine with different job orderings (due-date-first, shortest-job-first, longest-first, critical ratio, seeded shuffles, or every permutation when the job count is small), scores every complete schedule, and returns the best. It is guaranteed never worse than the baseline, and its badge says exactly what it did: "Best of N schedules tried."
• The exact solver encodes the schedule as a mathematical model, solves it, and returns a plan with a proven optimality gap: "proven within X% of optimal," and sometimes "proven optimal." Jobs using features the model does not yet optimize natively are locked to their current placement, so the solver can never emit an invalid plan.

Either way, nothing changes until you review the side-by-side comparison (late jobs, tardiness hours, makespan, how many operations moved) and click Accept. Discard leaves the database byte-identical. The full story, including the goal presets and the Explain dialog, is in the optimizer guide.

Quoting and What-If Promise Dates

When sales asks "can we ship 20 of these by Friday?", EDGEBIC answers with a simulation, not a guess. The quote module runs the identical scheduling engine against your real committed capacity, in memory, without writing anything to the live schedule. You get an earliest completion date, the cost rollup (hours per step times each work center's rate, plus material), and margin at your markup. Accept the quote and it converts to a manufacturing order with one click. Because simulation and live scheduling share the same engine code, the promise you quote is the schedule you get. The quote-to-ship walkthrough runs a titanium bracket order through the entire lifecycle, and the quoting guide covers scenarios and comparisons.

Inventory, Forecasting, and Make-to-Stock

For stocked products, EDGEBIC adds the planning layer above scheduling: a ledger-backed on-hand balance (every receipt and issue is an auditable ledger entry), safety stock and reorder points, demand forecasts, replenishment suggestions, and a Master Production Schedule grid where firming a bucket creates a build-to-stock manufacturing order. At scheduling time, the engine nets demand against stock: if a customer orders 80 and you have 60 on hand, the engine builds 20. A make-to-stock producer scheduled in the same run as its consumer is deliberately sequenced first, so the consumer sees the stock it will produce. The replenishment walkthrough and the consume-from-stock walkthrough show both halves with real numbers, and the planning guide covers the whole layer.

Insight: Gantt, Reports, and Dashboards

The Schedule tab is where you live day to day: the Drive Schedule grid for orders, the Job View Gantt with color-coded operation states and drag-to-adjust, the Schedule View, a Resource Calendar showing capacity and backlog per work center, and per-step detail views. Reports cover on-time delivery, utilization, earned value, and OEE, and every report column carries a built-in plain-language glossary (the Column Details button), so "SPI" is never a mystery abbreviation, and the wider vocabulary is collected in the EDGEBIC scheduling glossary index. Dashboards give KPI tiles and job-status views for the morning standup. The visual scheduling guide tours all of it.

Getting Started: From Installer to First Schedule

The path from zero to a working schedule is deliberately short:

1. Install and launch. Run the installer; on first launch EDGEBIC creates a local database file automatically. There is no database software to install for a single user.
2. Create the administrator and sign in. First launch asks you to create the admin account; you add named users and roles later under Settings, so audit history records who did what.
3. Follow the First Time Setup Wizard. Six steps in dependency order: data source, first shift, first product, work centers (at least three), a routing, and a first manufacturing order. Each step embeds the full editor, and you can stay on a step to add more records.
4. Or import instead of typing. Every master-data list (products, work centers, routing steps, orders) loads from Excel or CSV through import masks. This is also the ERP integration path: whatever your ERP can export, EDGEBIC can map and import.
5. Schedule and read the result. Open Schedule, run Drive Schedule, and check the Gantt in Job View.

Scaling to a team is a settings change, not a project: point every installation at a shared SQL Server database and all planners see the same schedule. The greenfield setup walkthrough traces the whole journey with a real bracket plant, and the admin guide covers users, security, and data sources.

What EDGEBIC Is Not

Honesty is cheaper than disappointment, so three boundaries worth stating plainly:

• It is not an ERP. No general ledger, no purchasing, no invoicing. It integrates with those systems through import-export masks, and the heritage integrations (the vendor-recommended Fourth Shift work, Macola, the Cummins AS400 deployment) belong to the User Solutions product line's track record.
• The optimizer is not magic. The multi-run layer is a guided search that is provably never worse than your baseline. The exact solver gives you a mathematical optimality gap. Neither claims "always optimal," and neither applies anything without your Accept.
• It will not fix bad data. Routings with wrong hours produce schedules with wrong dates, in EDGEBIC as in anything else. The difference is that EDGEBIC's diagnostics and troubleshooting tools make bad data visible fast.

Where to Go Next

If you are evaluating, the fastest way to judge EDGEBIC is against your own data: bring an ERP export of products, work centers, and routings to a demo and watch the first schedule come out. If you are learning, the library is organized for it: the engine deep dive for how scheduling decisions get made, the how-to library for step-by-step tasks, the worked examples for end-to-end scenarios with real numbers, and the results guide for what the outcomes look like in percentages and hours. And if you have been running production scheduling on general principles, the foundation post on what production scheduling is connects the generic discipline to what you have just read.

Contact US for a personalized demo and bring your messiest week of orders. That is the week EDGEBIC was built for.