P/N
Minimum
Location
Order
Supplier
For large manufacturers
Keep the small parts from becoming the reason a line stops. Arda puts a scannable reorder card where your teams use each item. Give the shop floor a direct way to signal what it needs while your ERP keeps running the plan.
P/N
10,000+
orders placed by a scan
$75M+
in parts
10,000+ orders, over $75M in parts, placed by a scan.
Arda platform to date
Never stop a line for something that was never on the plan
Every material that goes into the product is planned, purchased and reported against. The sealant, the blades, the gloves, the filters and the packaging at the line are not, because they were never on a bill of materials for anything to plan them from. Arda is inventory replenishment software that makes the reorder signal physical at the point of use. The card on the shelf carries the quantity, sized to real burn rate and your supplier's lead time, so replenishment starts the moment stock runs low, on one cell, one plant, or every site you decide to put cards on.
Your ERP is not wrong about the shelf. It was never asked about the shelf.
The plan is complete, and the shelf is still empty
A large plant does not have an information problem about its build. It has a complete, accurate, auditable picture of every material that was planned into the product. The gap starts at the pallet edge, on the items nobody planned because nothing exists to plan them from, and it is the same gap at every site. It is worth what it is worth: Siemens puts an hour of unplanned downtime in a large automotive plant at $2.3 million. (Siemens, The True Cost of Downtime 2024)
The more authoritative the system, the more invisible the gap
Corporate reporting is complete and accurate on everything that was planned, so a shortage at the line never surfaces as a system failure. It surfaces as somebody's bad morning, gets solved locally in twenty minutes, and is never visible above the plant. Nobody is being kept in the dark. The question was never asked.
Every plant already solved this, and no two solved it the same way
One site runs a two-bin rack a supervisor built out of scrap. One runs a shared sheet that one buyer maintains. One has a stockroom attendant who is, personally, the system. All three work, none of them roll up, and each one leaves with the person who built it. Corporate cannot compare four sites because there is nothing comparable to look at.
The fix that gets proposed is another project
The gap is real, so somebody scopes extending the ERP to the point of use, or a module, or a program with an integration attached. It gets costed. Then it goes into the same queue as the capacity work, and a consumables project does not beat a capacity project in any budget cycle that has ever happened. Two cycles later the line is still hunting for blades.
How does point-of-use replenishment work in a plant that already runs an ERP?
Arda's inventory replenishment software for large manufacturers puts a scannable card on the stock that sits where the work happens: the lineside racks, the cribs, the shadow boards, the supply cabinets. Whoever empties it scans it with the phone in their pocket, and the reorder fires at the agreed quantity to the supplier you already buy it from. Your ERP keeps planning, routing, costing and reporting exactly as it does today. Nothing about the corporate system changes, because nothing about the corporate system has to.
Land at one cell, then expand site by site
A plant does not need a rollout to find out whether cards hold. It needs one shelf, a fortnight, and a number it can hand the next plant manager. Each stage below is a decision by a real person, and none of them is a project. One cell: the ten items that stop the line most often, approved by the plant manager or CI lead, with corporate doing nothing since there is no hardware on the network, on the Free plan. One plant: every crib, lineside rack and supply cabinet in the building, approved by the plant manager with the site buyer once a supplier is onboarded and IT runs its review, on Growth then Pro. The second plant: the same card set resized to its own burn rate and suppliers, approved by the second plant manager looking at the first plant's scan history, with nothing new for corporate to approve, on Pro. Every site: ERP integration, custom workflows and one view across sites, approved by corporate operations and IT, the first stage where corporate does real work, on Enterprise. The corporate ERP is authoritative at every stage, including the last one.
- One cell
- One plant
- The second plant
- Every site
Card the stock at the point of use
Every item on the shelf gets a durable card with a QR code. An operator scans it with any phone when it runs low. No scanner gun to issue, no login to provision, no training day to schedule against the production plan.
The order fires automatically
The scan places the reorder at the sized quantity from the supplier you already buy it from. Nobody types a purchase order and nobody waits for a planning run to notice.
Pulled, on order, restocked, per cell and per site
The plant manager sees the whole building. The CI lead sees which cards move and which never do. When there are four sites, that same view rolls up across all four.
Each card carries a loop the plant's own standard work already recognises: the reorder point, the two-bin system, the safety stock that covers the lead time, and a pull system that fires on consumption instead of on a forecast. The arithmetic lives on the shelf rather than in a plan-for-every-part spreadsheet that one person maintains.
Hydraulic oil, ISO 46
Reorder qty
20 drums
Lead time
7 days
Supplier
Grainger Industrial
Safety stock
8 drums
Scanned 06:52
PO sent 06:52
Restocked
Start on the one line that stops most often
Card ten items on a single cell this week and let the scans come in. No hardware on the network, no integration to build first, and nothing for corporate to approve before you can see whether it holds.
Make Free Kanban CardsFree plan, no credit card
Manufacturers who already had systems, and still had the gap
These are independent shops, not multi-plant manufacturers, and this page is not going to pretend otherwise. What they share with your site is the part that matters: they owned software, the software did not reach the shelf, and putting a card at the point of use closed the gap without anybody starting a project. Same card, smaller building.
0
stockouts
20 days
order backlog, from 365+
Fat Fender
Parts manufacturer and custom vehicle shop, after two failed ERP implementations. It is the most useful proof on this page for one reason: the objection your own team will raise is that the last rollout did not stick, and the card motion held exactly where heavier software had not.
Read the Fat Fender case study →2x
throughput
10 min
a day ordering, from 2–3 hours a week
Assembly Industrial
Heavy-industry fabricator of custom trucks and metalwork. A manager used to walk the floor every week to work out what needed reordering. Project manager Anthony Olivas: "I just take 10 minutes at the end of the day to go around the shop, collect the cards, and now I know exactly what needs to be ordered. It used to take two or three hours a week."
Read the Assembly Industrial case study →100%
of supplies carded
1 yr
to full coverage
Rossmonster
Overland vehicle builder. Cards went on 100% of the supplies they stock inside a year, one shelf at a time, with no implementation project. That is the expansion pattern this page is describing, run to completion at one site.
Read the Rossmonster case study →Three ways a plant closes the point-of-use gap
Nobody at a large manufacturer is choosing between Arda and their ERP, and this page would be useless if it pretended they were. The real choice is between funding a project to extend the system you have, bringing in a supplier-run program, and putting a card on the shelf this week.
ArdaExtending the ERP to the point of useA supplier-run vending or bin program
What fires the reorderA scan on the card at the shelf, by whoever emptied itA transaction someone posts, then a planning runThe supplier's round, or a dispensing event at a machine
Who approves the first siteThe plant. Nothing goes on the networkA funded project, plus IT and the system ownerProcurement, an agreement, and usually a site-volume floor
Time to the first live shelfAn afternoonA budget cycleWeeks, plus hardware to install
What it reachesEverything you stock at the point of use, on-BOM and offWhatever gets configured, which starts from the planned material it already holdsThe lines the program covers, at the sites that justify a stop
Adding the second plantCard it. There is nothing to implement againPhase twoA new site agreement
Effect on the corporate ERPNone. It stays authoritative and unchangedIt is the ERP, and it is the right answer for planned materialNone, though the consumption history sits with the supplier
Cost to start$0 on the free planProject budgetBundled into unit price
Frequently asked questions
Does Arda replace our ERP?
No, and it is not built to. Arda runs one loop, replenishment at the point of use, on the shelf stock that never made it onto a bill of materials for your ERP to plan from. Your ERP stays the system of record for jobs, routing, costing, planned material and financials, and none of that has to move for the first card to work. If you run SAP, Oracle, Infor or NetSuite, keep running it. Arda is not an alternative to any of them and does not want the job they do. It covers the shelf they were never pointed at.
Can we run this at one plant without a corporate rollout?
Yes, and that is the intended way in. Start with ten cards on one cell, on the Free plan, with no hardware on your network and nothing installed on a machine. A plant manager or CI lead can put that in place inside a week without a budget line or a project. Once the cards are moving, the case for the rest of the building is your own consumption history rather than a vendor's slide, and the case for the second plant is the first plant's numbers. Corporate does real work only at the point where you want ERP integration and a single agreement across sites, which is the Enterprise plan and is the last step rather than the first.
What will IT, security and procurement need before we can pilot?
Less than they expect, because of what Arda does not do. There is no appliance on your plant network, no agent installed on a workstation, and no connection to a PLC, a line controller or a historian, so nothing touches the OT side of the house and nothing crosses the IT and OT boundary your reviewers protect. The scan happens on a phone against a printed card, and the data involved is item names, quantities, suppliers and scan times rather than personal or production data. Integration is optional and comes later: Arda connects to QuickBooks, NetSuite, Odoo, Xero, Shopify, Stripe, Airtable, Make and Zapier, plus a REST API, and until you turn one on there is nothing pointed at your systems at all. For a formal vendor security review, a completed questionnaire, or anything your team needs on paper before a purchase order, talk to us about the Enterprise plan and we will work to whatever your process requires.
Is a kanban card the right method for a plant that already runs planning at this level?
Kanban was not invented for small shops. Toyota built it as the signalling mechanism of the Toyota Production System precisely because a plan issued centrally cannot see what a station has actually consumed, and it spread from the plants to the suppliers for the same reason. Nothing about running a full ERP changes that logic. Planning is still the right answer for the material the plan can see, and a physical pull signal is still the right answer for the material at the point of use, which is the class your value-stream map has always drawn as an assumption. What Arda adds is that the card also places the order, so the loop closes without anyone posting a transaction or maintaining a plan-for-every-part sheet by hand.
How does this get governed once it is running at five sites?
One supplier record, one agreement, and one view across every site. Each plant sizes its own cards against its own burn rate and its own suppliers, because a card sized for a plant running two shifts is wrong at a plant running one, and every scan is timestamped to the card, the item and the site. Pro adds AI-tuned card quantities and multi-card loops for the items that need more than two bins. Enterprise adds ERP integration, custom workflows for lineside delivery, and the roll-up corporate needs when this stops being one plant's idea. Scanning is free and never takes a seat, so there is no licence position to true up as the fifth site comes on.
How much does Arda cost, and is there a free trial?
Create unlimited items and cards on Free, without Order Queue access. Solo includes one user and up to 10 orders a month after its 7-day trial. Team and Pro are priced per Power User and include unlimited scanning for your team. Compare current plans and billing.
Ten cards, one cell, and nothing for corporate to approve
Pick the line that stops most often and card the ten items that stop it. No hardware on your network, no integration to build first, and no change to the system your company runs on. Prove it on one cell, then hand the next plant the numbers.
Free plan, unlimited cards, no rollout project.