Machine shop production planning & scheduling

What machine shop planning means

Machine shop planning is material requirements planning (MRP) and production planning and control (PPC) built around finite machine loading and Gantt scheduling, for a high-mix, low-volume shop that shares a set of CNC, VMC, turning and grinding machines across many jobs at once. It takes each confirmed order, explodes its bill of materials, nets against stock, and raises work orders that each carry a routing — the sequence of operations, each on a machine, each with a standard cycle and setting time. It then loads each operation onto its machine, shows the percentage load on every resource, and sequences the shared queue by order and resource priority on a Gantt board.

Where a flow line runs one product down a dedicated path, a job shop is the opposite: dozens of different parts competing for the same handful of machines. So the central question is not "what to make" — MRP answers that quickly — but "in what order, on which machine, to keep everything busy and still hit the dates." Machine shop planning is, at heart, a scheduling problem wrapped around an MRP core.

This guide pairs with the machine shop planning software product page, which shows the screens; here we explain the method that makes shared machines behave.

Why a job shop is a scheduling problem

Three characteristics make a machine shop distinct from a repetitive plant, and all three point at scheduling.

1. Many jobs, few machines, one queue

The same VMC or cylindrical grinder is wanted by a dozen orders in the same week. Every job's route crosses several shared machines, so a delay on one bottleneck ripples through everything queued behind it. Planning has to see each machine as a contested resource with a queue, not as infinite capacity.

2. The quote depends on the schedule

In a job shop the delivery you promise at quotation is a scheduling decision. If you promise from a gut feel, you either lose the order to a shorter lead time or win it and miss it. Knowing each machine's projected availability before you quote is what lets you commit to a date you can actually keep.

3. Setup time is real money

Every changeover costs setting time on an expensive machine. The sequence you choose — which jobs run back-to-back, which similar setups you batch together — directly changes how much capacity is left for productive cutting. A good schedule is not just on-time; it is setup-aware.

Routing, standard times and the work order

Everything downstream depends on one thing being right: the routing. Each work order carries a process sheet — an ordered list of operations (turning, milling, drilling, grinding, inspection), each assigned to a machine or work centre, each with a standard cycle time per piece and a setting time. That is the data finite loading and scheduling consume. Get the standard times roughly right and the whole plan is trustworthy; get them wrong and every load figure and every promise date inherits the error.

The order still begins as demand. A confirmed job is entered as a sales or production plan, exploded through its BOM, and netted against stock and open supply so the raw material and bought-out shortfalls become purchase requisitions while the parts you machine become work orders. Each work order's routing — defined per part or per category and printed as a job card individually or in batch — is what turns "make 200 of this" into a set of timed operations that can be loaded onto real machines. See Process Sheets & Routing.

Finite machine loading, explained

Once operations are routed, finite machine loading compares the work queued on each machine against the hours it actually has.

• Under 100% — the machine has projected availability; it can absorb more or pull a job forward.
• Around 100% — the machine is fully committed; there is no slack for a rush order.
• Over 100% — the machine is the bottleneck; re-sequence, move work to a similar machine, or add a shift, or the date slips.

The crucial word is finite. Infinite loading assumes a machine can absorb any amount of work — useful for a rough view, dangerous for a real week, because it hides overloads until a job is already late. Finite loading refuses to promise more than the shop can run, so an overload surfaces as a number above 100% while there is still time to level it. See Machine Loading & Capacity and the worked examples in machine capacity planning explained.

Sequencing shared machines on a Gantt

Loading tells you whether the week fits; sequencing decides the order in which the queue runs. Because a job shop's machines are shared, scheduling is about arranging a contested queue rather than dedicating a line. Fast Planning sets and re-sequences work-order priority and lays the operations out on a DayPilot Gantt board, so a planner can see each machine's queue at a glance, drag a rush job forward, and level work across machines and shifts before a date slips.

Two ideas make sequencing productive rather than reactive. Priority — by order and by resource — lets the genuinely urgent job move without the loudest customer simply jumping the queue. Projected availability — when a busy machine will next be free — is what lets the shop promise the next job a realistic date instead of an optimistic one. Sequenced well, similar setups can be grouped to cut changeover time, turning the schedule itself into a lever on capacity. See Scheduling & Priority (Gantt) and the deeper walk-through in the production scheduling guide.

Inward and outward job work (ITC-04)

A machine shop is usually on both sides of job work. Work you do on a customer's material — inward job work — is planned and loaded like any other work order, so it competes for machine capacity on the same board as your own parts; it just consumes the customer's material rather than yours. Operations you send out — heat treatment, hard chrome, special grinding — are planned as OSL (out-sourced / outside-labour) components, raised as work orders routed to an outside processor and tracked out and back, so an outside operation running late shows up in the plan as a bottleneck rather than a surprise at assembly.

For Indian shops, both the material you receive to work on and the material you send out are movements under the GST job-work provisions, and the quarterly ITC-04 return summarises inputs sent to and received from a job worker. Tracking those movements inside the plan is what makes the ITC-04 reconcilable against reality rather than reconstructed at quarter-end. Filing thresholds and treatment change from time to time, so confirm the current position with your CA. The planning payoff is constant: the tax movement and the capacity plan read from the same records, not two disconnected registers.

Machine utilisation and OEE

The reason to plan finitely is to run the shop close to its real capacity — and you can only manage what you measure. As the shop runs, operators book progress against each operation by scanning shift, machine and operator, recording actual start and end, setting and cycle times, stoppages and OK-versus-not-OK quantity. That booked data drives the numbers that tell a shop where its hours actually go:

Seeing which machines run below their benchmark, and why, is what lets a shop attack idle and setting time, re-sequence to cut changeovers, and load closer to true capacity. Because every factor is measured from booked reality, not estimated, the numbers stand up in a review and feed the next schedule. Barcode and machine-data capture make this practical at shop speed — see IoT / Industry 4.0 machine data, Plan vs Actual & OEE and OEE & efficiency tracking.

How Fast Planning implements it

Fast Planning Software for machine shops is a working implementation of everything above, built by Improsys in Pune under the Fast Technology brand, deployable cloud or on-premise. Mapping the cycle to the product:

Because it runs on the shared platform, the same deployment hands work orders to Fast Production, reads and reserves stock in Fast Inventory, and feeds purchase requisitions into purchasing — with no interface and no re-keying. Schedule releases and reorder warnings can push to the shop and to suppliers over WhatsApp, email and SMS.

Deployment, India context and pricing

Fast Planning suits the small-to-mid Indian machine shops that fill the Pune, Chennai, Rajkot, Coimbatore, Ludhiana and NCR engineering clusters — single-plant CNC and job-work units running a shared bank of machines. It deploys cloud or on-premise on IIS and SQL Server and is licensed as a branded single-tenant copy, so your routings, standard times and machine-hour rates stay in your own database.

Two India realities shape the plan. First, job work is the business model for many shops — material received to work on, and operations sent out — so inward-and-outward tracking and ITC-04 reconciliation belong inside the plan; confirm the current tax treatment with your CA. Second, machine-hour costing is the margin: because your quote rests on standard cycle times and a machine-hour rate, measuring real utilisation and efficiency against those standards is what protects the margin you priced. Pricing is indicative in Indian Rupees and scales with users and modules; the practical step is a quote against your machine list and routings, and a short demo on your own jobs. See planning software pricing or talk to us.

Frequently asked questions