In a steel plant, a missing spare rarely costs the price of the part. It costs the hours a line sits waiting for it. Most storerooms still reorder with min and max levels set years ago, while usage of bearings, rolls, electrodes and caster nozzles moves with production, PM plans and equipment condition. The result: slow movers on the shelf and a stockout on the part that matters. Autonomous AI agents forecast spares usage from real plant plans and raise requisitions before stockouts hit, with buyers approving every order. Start with a spares review.
Steel Plant Spare Parts Reordering with Autonomous AI Agents
An inventory agent reads PM plans, failure predictions and production schedules, forecasts what each critical spare will need, and raises purchase requisitions in your ERP for buyer approval.
- Why fixed min and max levels fail in steel plants
- Forecasting spares from PM plans and predictions
- Requisitions, buyer approval and voice stock checks
Why Fixed Min and Max Levels Fail
Steel plant spares do not move at a steady rate.
Minimum and maximum levels assume demand is roughly steady. In a steel plant it is not. Caster nozzle use rises with sequence changes, electrode use follows tonnage and practice, bearings are used in bursts during outages, and some critical spares sit for years before they are needed once. Fixed levels set long ago cannot follow these patterns. Our MRO specialists can review how your levels were set.
Planned use ignored
PM and outage plans not fed into reorder levels.
Lead times change
Supplier lead times move but levels do not.
Slow movers pile up
Money tied up in parts nobody uses.
Critical stockouts
The one part that matters is not on the shelf.
Different Spares Need Different Rules
One reorder method for every part is the root of the problem.
A good inventory policy groups spares by how they are used. Consumables follow production. Planned replacements follow PM plans. Repairable parts cycle through workshops. Insurance spares are held for rare, high-impact failures. The agent applies the right logic to each group, and stores teams can move a part between groups when its role changes. Classify your spares with our inventory team.
A spare transformer bushing or a mill motor may not be used for years, so a usage-based forecast would say to hold none. Insurance spares are held because of the cost and lead time of the failure they cover. Review them each year with reliability and finance, separately from routine reordering.
Forecasting From Plans and Predictions
The best signal for future use is what the plant already plans to do.
The agent builds each forecast from several sources at once: past usage, the production plan, PM and outage schedules, and failure predictions from condition monitoring. A planned outage that replaces forty bearings shows up in the forecast weeks ahead. A rising vibration trend on a gearbox raises the chance its spares will be needed. Each forecast shows which source moved it, so stores and buyers can see why a number changed instead of trusting a black box. Map your data sources in a data session.
Melt and cast plan
Tonnage, sequences and grades drive consumables.
PM and outage plans
Parts lists from planned jobs.
Failure predictions
Rising risk on assets brings spares forward.
Lead times
Actual supplier delivery times, not catalogue values.
A Worked Reorder Point Example
When to order, explained in one line of maths.
A simple reorder point is the expected use during the supplier's lead time plus a safety stock for uncertainty. The agent keeps the same structure but feeds it with a forecast instead of a flat average. Here is an illustrative example for caster SEN nozzles. Run your own numbers in a working session.
Order Critical Spares Before They Run Out
Forecasts built from your production, PM and condition data, with requisitions raised in your ERP for buyers to approve.
Spares for Planned Outages
The biggest bursts of demand are the easiest to see coming.
Outages and relines use large numbers of bearings, seals, wear parts and refractories in a few days. When outage job lists are built early, their parts lists feed straight into the forecast. The agent compares them with stock and lead times weeks ahead and flags anything that will not arrive in time, so the outage is not delayed by a missing kit. Link outage plans with our planning team.
Parts list checked
Every job's parts compared with stock and lead time.
Late items flagged
Parts that cannot arrive in time raised early.
Kits reserved
Parts reserved by job so they are not used elsewhere.
Slow Movers and Excess Stock
Better forecasting also shows what you should stop buying.
Every storeroom has parts that have not moved for years: spares for equipment that was replaced, duplicates under different part numbers, and quantities bought for a project long finished. The agent flags these by comparing stock with forecast use and with the equipment still installed. Stores and reliability teams then decide what to keep as insurance, what to use up and what to sell or scrap. Review slow movers with our inventory team.
What gets flagged
- No issues for a long period
- Parts for equipment no longer installed
- Likely duplicates under other numbers
- Stock far above forecast use
What people decide
- Keep as an insurance spare
- Use up before reordering
- Merge duplicate part numbers
- Return, sell or scrap
Requisitions, Not Surprise Orders
The agent prepares the order. A buyer places it.
When stock is forecast to fall below its reorder point, the agent drafts a purchase requisition in your ERP with the quantity, the reason and the forecast behind it. Buyers review, adjust and approve as usual, and keep control of suppliers, prices and contracts. Where a contract or framework agreement exists, the requisition can point to it, so routine reorders move quickly while unusual quantities still get a closer look. Set approval rules with our procurement team.
Forecast
Use predicted from plans and condition.
Check
On hand, on order and lead time.
Draft
Requisition with quantity and reason.
Approve
Buyer reviews and places the order.
Track
Delivery followed against need date.
Voice Stock Checks From the Floor
Ask the storeroom without leaving the job.
Technicians and supervisors often need to know whether a part is in stock before they plan a job. With voice, they ask from the floor and get an answer straight away, including where the part is and whether any are reserved. Ask our voice team about hardware.
Repairable Spares and Rolls
Some spares come back. Track the whole pool.
Rolls, segment rollers, pumps and gearboxes are often repaired or reground rather than bought new. The question is not how many to buy, but whether the pool of parts in use, in repair and on the shelf is big enough for the change rate and repair time. The agent tracks each pool and warns when repair delays or regrind limits threaten availability. Discuss repairable pools with our reliability team.
What the agent tracks
- Parts in use, in repair and ready
- Repair turnaround by workshop
- Regrind count and remaining life of rolls
- Upcoming change dates from the plan
What it warns about
- Ready stock below planned changes
- Repairs running late
- Rolls near their scrap diameter
- Pools that need a new purchase
Spare Parts Reordering Mistakes to Avoid
Automation on bad master data just orders the wrong parts faster.
These are common problems when plants automate spares reordering. A short readiness review can find them.
What works
- Clean part numbers and descriptions
- Spares grouped by how they are used
- Real supplier lead times
- Buyer approval on every requisition
What fails
- Duplicate parts under different numbers
- Insurance spares forecast from usage
- Ignoring PM and outage parts lists
- Orders placed with no human review
How iFactory Runs Spares Reordering
Autonomous Inventory Agent, on your own network.
iFactory's inventory agent forecasts critical spares from PM plans, production schedules and failure predictions, raises purchase requisitions to your ERP for buyer approval and answers stock questions by voice. It runs on an on-prem AI server inside your plant. Fit questions go to our solutions team.
Spares forecasting
From production, PM and condition data.
Reorder points
Refreshed as plans and lead times change.
ERP requisitions
Drafted with quantity and reason.
Buyer approval
Every order reviewed by a person.
Voice stock checks
Answers from the floor in seconds.
Repairable pools
Rolls and repairables tracked through repair.
Results depend on your master data, suppliers and plans. We measure stockouts, slow movers and requisition approval rates on your own data during the pilot, rather than promising a general figure.
Turnkey AI: Delivered, Connected and Live in 6–12 Weeks
You do not build this. It arrives ready.
iFactory ships as a pre-configured NVIDIA AI server with the software pre-loaded, racked and ready. Rack it, plug in power and Ethernet, and the AI is live on your network. Our team handles cabling, network setup, PLC, SCADA, DCS and CMMS integration, ERP and CMMS links, operator training and 24×7 remote monitoring. Data stays on your network. For a scope matched to your plant, request a turnkey quote.
Ship, network and data
Server installed. ERP stock, CMMS PM plans and production schedule connected.
Model training and pilot
Spares grouped and forecasts tuned. Requisitions drafted for buyer review in one area.
Go-live and training
Inventory agent live for critical spares. Buyers and stores trained. 24×7 monitoring begins.
Frequently Asked Questions
What is autonomous spare parts reordering?
An AI agent that forecasts spares use from plant plans and condition data, checks stock and lead times, and drafts purchase requisitions before stock runs out.
Does the agent place orders by itself?
No. It drafts requisitions in your ERP. Buyers review, adjust and approve them, and keep control of suppliers and prices.
Which spares benefit most?
Critical consumables and planned replacements with variable demand, such as caster nozzles, electrodes, bearings and seals.
How are insurance spares handled?
Separately. They are held by risk and reviewed yearly, not forecast from usage.
How does it use failure predictions?
When condition monitoring shows rising risk on an asset, the forecast for that asset's spares rises, so parts are ordered before they are urgently needed.
Can it track rolls and repairables?
Yes. It tracks the pool in use, in repair and ready, along with repair turnaround and regrind limits.
Can it help reduce excess stock?
Yes. It flags slow movers, likely duplicates and parts for equipment no longer installed, so stores and reliability teams can decide what to keep, use up or remove.
Can I ask about stock by voice?
Yes. Questions such as "how many SEN nozzles in stock?" are answered with quantity, location and reservations.
How do we start?
Start with one area's critical spares, such as the caster. A 6-week pilot connects your ERP and plans and drafts requisitions for buyer review. To plan it, contact our team.
Never Wait on a Critical Spare Again
In thirty minutes we review your critical spares list, reorder levels and recent stockouts, and show where forecasts would change your orders. You keep the notes either way.
- 1Your critical spares list
- 2A year of issue history for those parts
- 3Current min and max levels
- 4Supplier lead times
- 5The last stockout that stopped a line
