Where the master plan and the floor drift apart
PPC turns the order book into loom and line loading across spinning, weaving or knitting, processing and stitching. The trouble starts when the plan, the floor and the order book live in different places.
The master plan lives in one planner’s workbook
The master plan, the loom and line allocation chart and the capacity model sit in a planner’s Excel file, often in more than one copy. Each unit keeps its own. When the plant head asks what line 7 runs next week, the answer depends on who opens which file. The playbook warns that the first time a single master-plan file is corrupted, the team can lose two days finding the version the planner was using.
32%
Share of about 900 Bangladeshi manufacturing firms, surveyed in 2019–20, that plan production in general spreadsheets such as Excel rather than planning software. 53% still plan by hand and 10% use ERP. Five industries, garments among them.
Source: World Bank, Gearing Up for the Future of Manufacturing in Bangladesh (opens in a new tab), 2021.
Plan versus actual, found out at the end of the week
Supervisors write hourly output on a board or plan card, and it reaches a sheet days later. In a spinning or weaving mill the same gap opens in kilograms per shift or metres per loom, and shows at month end. The playbook says a line 20% behind by 11 am will not recover that day. Nobody can say quickly where the gap opened, or whether the cause was an input, such as an absent operator or short fabric, or rework.
70%
Share of its planned output that one Ethiopian spinning mill produced over 12 months: 1,400 t against 2,000 t. Its on-time in-full delivery rate was 62.3%. A single-mill case study.
Source: Textile Science & Research Journal, Assessing the effectiveness of the production planning system in spinning mills: A case study at Bahir Dar Textile Share Company (opens in a new tab), 2026.
“Can it still make the ship date?” answered by feel
Merchandising asks whether an order can leave a week early. The answer is arithmetic: line-days or loom-days still needed against those free before ex-factory, if fabric and trims arrive on time. The figures sit in the ERP, the SMV list, the capacity sheet and the inbound dates, so the answer comes days late. The playbook warns: a buyer accepts a later ship date if told early, but not on the ship date itself.
A breakdown or a late yarn lot, re-planned by phone
A loom goes down, a yarn lot arrives late, a shade is rejected, or the buyer swaps two orders. The planner must find what that resource was carrying, which ship dates move and which option fits: absorb, compress downstream, push with the buyer’s consent, or split the order. The playbook asks for the new plan within four hours, with the replan logged. When the re-plan is agreed by phone, nobody fills in the log.
Five planning jobs, worked from your master plan and order book
It reads your master plan, allocation chart, capacity sheet, plan-versus-actual rows and Odoo production orders. Every answer cites its rows and records. A planned-start change, a log row or a message to a unit head waits for an admin to approve or decline.
Ship-date checks in the playbook’s green, amber and red
Answers with sources
Upload the master plan and a capacity sheet, one row per line and week, as Excel or CSV. The agent reads the order in Odoo, with the time fetched, then calculates the line-days needed from the plan and the free line-days for each line and week you name from the capacity sheet, citing the rows. It holds the totals to the playbook’s rule: green with 10% slack or more, amber if it fits with none, red if it does not.
Ask“Can order 2231 still make its 21 March ex-factory on lines 4, 5 and 6? Show the line-days needed, the free line-days in each week up to then, and the rows used.”
Free capacity next month, by stage, from your capacity sheet
Answers with sources
Upload the capacity model as Excel or CSV, one row per unit, stage and month, with committed and free loom-days or line-days. For the month and stage you name, the agent adds up both, works out utilisation per unit, marks any unit over the playbook’s 95% and lists the rows. Dyeing batch slots work the same way. Loom efficiency % comes from an uploaded export or a Google Sheet, not the ERP.
Ask“How many loom-days and line-days are uncommitted in December, by stage? Which units are over 95% utilised?”
A re-plan proposal with every assumption written down
Creates files
After a breakdown, yarn delay or buyer change, ask for a re-plan proposal. The agent writes it as a document: the orders and lines affected, their ship dates, and the playbook’s four options (absorb within slack, compress downstream, push with the buyer’s consent, split across lines), each with the recorded slack and line-days it rests on, and what the records leave out. The planner chooses. Every revision is kept.
Ask“Loom 14 is down until Thursday. Draft a re-plan proposal for the orders it was carrying and state every assumption you used.”
A new planned start in Odoo, waiting for an admin
Reads your ERP
Once the planner settles the re-plan, the agent can draft the change to the production order in Odoo: a new planned start, and who it is assigned to. A department admin approves or declines it, and it runs only if approved. It does not create orders or change quantities, and it writes nothing to any other ERP.
Ask“Move the planned start of production order MO-0412 to Thursday, as agreed in this morning’s re-plan.”
The replan log row, and the new loading to the unit head
Acts in your tools
The playbook wants every replan’s reason, time, decision and owner in a log. The agent drafts that row for your replan log in Google Sheets, and the new loading as a Slack message to the unit’s channel or a WhatsApp Business message to the unit head. An admin approves or declines each one, and nothing is written or sent unless approved.
Ask“Log today’s re-plan for order 2231 in the replan sheet, and send the new loading for line 6 to the sewing unit head on WhatsApp.”
A week at the PPC desk, in three asks
Merchandising asks to pull an order forward
Master planner
The buyer wants order 2231 a week early, and merchandising asks planning whether it can leave on the 14th.
Asks
“Can order 2231 ship on the 14th instead of the 21st? Show the line-days it still needs on lines 4, 5 and 6, what is free in the weeks up to the 14th, and which orders share those lines.”
The agent
Reads the production order in Odoo, with the time fetched. Calculates the line-days still needed from your master plan, and the free line-days on lines 4, 5 and 6 for each week up to the 14th from your capacity sheet, one figure per line and week, citing the rows. Holds the totals to the playbook’s 10% slack rule for a green, amber or red. Lists the other orders loaded on those lines, each with its ship date and recorded slack, and cites every row.
Result
A green, amber or red answer with the figures behind it, and the orders sharing those lines, so the planner can put the cost in days to merchandising and the plant head, as the playbook asks.
Loom 14 goes down mid-week
Weaving planner
Maintenance reports loom 14 down until Thursday. The plant head agrees to cover its remaining metres on looms 9 and 11, which already run the same quality.
Asks
“List what loom 14 was carrying this week, with the ship dates. Then log the re-plan and tell the weaving unit head that the remaining metres of order W-0587 move to looms 9 and 11.”
The agent
Reads the allocation chart in your connected Google Sheet and the production orders in Odoo. Lists each order on loom 14 with its ship date from the chart and its quantity from Odoo, citing the cells and the fetch time. Drafts a row for the replan log with the reason, time, decision and owner, and a WhatsApp Business message to the weaving unit head with the new loading.
Approval
Both wait as actions. The PPC head, as department admin, approves or declines each one. The row is written and the message sent only if approved. Each decision is kept in the audit trail under the PPC head’s name.
Result
Each runs only if approved: the row with the reason and owner is added to the replan log, and the message with the new loading is sent from your connected WhatsApp Business account. A declined one does not run.
The Monday plan-adherence review
PPC manager
Monday morning, before the weekly production meeting with the plant head.
Asks
“From last week’s plan-versus-actual sheet, give plan adherence by line and unit. Mark lines under 80% and the day each first fell 15% behind, as a spreadsheet.”
The agent
Sums planned and actual output from the rows of the uploaded sheet for each line and unit, works out adherence from those sums, and marks the lines under the playbook’s 80%. For each of those, it goes day by day to the first shortfall over 15% and quotes the reason column where one was recorded. It builds the result as a spreadsheet and lists the rows used.
Result
One sheet for the meeting, each figure traceable to its rows, and the lines with no recorded reason named, so the input-or-rework question gets asked there.
The cost of idle capacity and slipped stages, from published studies
These are published figures on capacity booking, buyer forecasts and stage delays in textile and garment supply chains, cited at the source, and not Mavick results. A pilot baselines your plan adherence, utilisation and replan frequency, then reads them again.
Free capacity known when a booking falls short
Over 36%
Share of suppliers who said gaps between the capacity a buyer booked and the orders it placed left capacity unused. 9.1% had to take last-minute, low-price orders to fill it. Apparel, footwear and household textiles, 2018.
How the agent helps
When a booked order comes in short or an RFQ arrives, ask what is free next month by stage. The agent adds up committed and free loom-days or line-days from the capacity table you uploaded, works out utilisation per unit and shows the rows. It checks the RFQ against the playbook’s rule for a tight month. It does not forecast demand or decide on the order.
Track: Capacity utilisation % by unit and month
Source: Better Buying, Better Buying Index Report, Fall 2018: Purchasing Practices Performance in Apparel, Footwear, and Household Textile Supply Chains (opens in a new tab), 2018.
Short-notice re-plans written down with their reason and owner
30.4%
Share of suppliers receiving buyer forecasts who got them less than 60 days ahead, in 2025 ratings by soft-goods suppliers. Late word of forecast changes can leave lines idle or force subcontracting, the report says.
How the agent helps
When an order changes, or merchandising passes on a new forecast, ask what it touches. The agent lists the affected orders and lines from your allocation chart and Odoo production orders, writes the re-plan proposal with its assumptions, and drafts the replan-log row. An admin approves or declines the row; it is written only if approved, so each replan’s reason stays on record.
Track: Replan frequency per week, by reason
Source: Cascale (Better Buying), Better Buying Purchasing Practices Index (BBPPI) 2025 report (opens in a new tab), 2025.
Stage delays traced by order, when you ask
Over 80%
Share of the lead-time delay on one 5,000-piece knit T-shirt order that came from five stages, among them lab dip, dyeing and sewing, in a Bangladeshi university case study. Every T&A operation finished 1–3 days late.
How the agent helps
Small slips at several stages add up. Upload your T&A or WIP sheet, one row per order and stage with a days-late column, and ask where an order lost time. The agent sums the days late by stage from the rows, lists them with the order’s recorded slack and shows the rows. A note to merchandising waits for an admin to approve or decline; it is sent only if approved.
Track: Critical-path slack on the top three orders
Source: BAUST Journal (Bangladesh Army University of Science and Technology), Case Study Investigation on the Hindrance Causing Delays in Time and Action (TNA) Calendar: Effects on Lead Time in Garments Industry (opens in a new tab), 2026.
What to baseline before a pilot
Measure each one before the pilot starts, then again at its end.
| KPI to baseline | Why it matters |
|---|---|
| Plan adherence, actual against planned output | The playbook targets 90% or more. Below 80%, it reads the plan as unrealistic or the floor as unstable. |
| Capacity utilisation % by unit and month | The number to check before you accept an RFQ in a tight month. The playbook’s band is 80–90%; above 95% leaves no slack to recover a slip. |
| Replan frequency per week, by reason | The playbook tracks it as a trend. A spike points to unstable yarn or fabric supply, or to a plan that was unrealistic from the start. |
| Critical-path slack on the top three orders | The playbook says never to let it reach zero. At zero on a top buyer’s order, the plant head and the merchandising head decide within four hours whether to push the ship date or compress. |
| Line balance index | Sum of station SMVs ÷ (stations × bottleneck SMV). The playbook targets 85% or more. A line below that misses its hourly target however it is loaded, so the playbook refers it to Industrial Engineering. |
What the agent works from
Records it reads
- Master plan
- Loom and line allocation chart
- Daily plan versus actual
- Capacity model
- Daily plan card
- Replan log
In your ERP
Reads Odoo production orders (order, product, quantity, status) and stock transfers (partner, status, scheduled date, source document). Drafts an Odoo planned start or assignee, or a transfer note, for an admin to approve or decline; it runs only if approved. Loom efficiency and ship dates come from your plan, an Excel or CSV export or a Google Sheet.
One ERP connection serves the whole factory: SAP Business One, SAP S/4HANA, Odoo, Oracle NetSuite, Microsoft Dynamics 365, WFX Cloud ERP or Datatex NOW. This agent reads only the records named above; we check yours on the call.
In your tools
Reads the master plan or capacity sheet you keep in Google Sheets when you ask. Prepares a row for the replan log in Google Sheets, the new loading as a Slack or WhatsApp Business message to a unit head, and an email to merchandising and the plant head from Gmail or Outlook. An admin approves or declines each one, and it runs only if approved.
Tools named today: Gmail, Outlook, Google Sheets, Slack and WhatsApp Business.
Never made up
- Capacity figures, machine counts or available hours
- Order quantities, ship dates or allocation decisions
- Production output for a shift or a day
- Lead times for any process stage
If it is not in your connected data, the agent says so.
Belongs to another department
- Buyer commitments and order changesMerchandising
- Why a machine is down, and when it will be repairedMaintenance & Engineering
- SMV, line balancing and capacity studiesIndustrial Engineering
- When suppliers commit to deliver yarn, fabric and other raw materialsSourcing & Procurement
What the Planning & PPC playbook covers
SOP: order acceptance in green, amber or red
Take the PO and tech pack from merchandising. Required line-days are SMV × quantity ÷ available line minutes per day. Check them against the master plan window, allowing for fabric and trim lead times. Green fits with 10% slack or more, amber fits with none, red does not fit without a push or a reallocation. Record the answer and the reason in the master plan.
Decision rule: a line 20% behind mid-shift
Two questions. Can the gap be recovered this shift, or does it carry to the next? Is the cause an input, such as an absent operator, short fabric or a machine down, or an output, such as DHU rework? Inputs can be fixed now. Outputs need a quality huddle, not a schedule fix.
KPI: capacity utilisation between 80% and 90%
Above 95% leaves no slack to recover; below 70% means too few orders or the wrong line balance. If utilisation stays over 95% for four weeks, the playbook takes it to the plant head and the COO: the plant is over-loaded, so decline new orders or add capacity.
Edge case: two orders, one dyeing recipe, one batch slot
Two orders share a dyeing recipe, and only one fits in the next available batch. The playbook gives the batch to the order with the earlier ship date, then recomputes the critical path of the order that was bumped. In woven fabric, it says, recipes are the real critical path: a yarn slip can be absorbed, a recipe slip cannot.
Written as general textile practice and reviewed by four independent AI models. Not yet signed off by a human subject-matter expert. Upload your own SOPs and standards, and every answer shows which source it used.
Ten questions it ships with
The template’s starter questions. Your team can edit them, or ask in its own words.
Can this order still make its ship date
Given the current stage and remaining capacity, can this order still make its ship date? Show the calculation and where the risk sits.
This week’s bottleneck
Which process stage is the bottleneck this week, what is it holding up, and how much output is being lost to it?
Plan versus actual by unit
Show plan versus actual output per unit for the last week, and tell me where the gap first opened.
Re-plan after a machine breakdown
A machine is down. Show me what it was allocated to, which orders are affected, and what the options are for re-allocating that work from current capacity.
Loom or line allocation for an order
Which looms or lines is this order allocated to, and what would be displaced if it were moved forward?
Capacity free next month
What capacity is uncommitted next month by process stage, at our current efficiency levels?
Orders at risk this week
Which orders in the current plan are at risk of missing their ship date? Rank them by how much time is left and say what is holding each one.
Impact of a yarn delay
If the yarn for this order arrives a week late, what does the rest of the plan look like? Show which stages shift and what they push out.
WIP position across stages
Show the current work-in-progress by order and process stage, and flag anything that has not moved for more than a few days.
Draft a re-plan proposal
Draft a re-plan proposal for the coming week from the live order book and capacity on record. State every assumption you used.
Planning & PPC: questions before a demo
Can it build or optimise our production schedule?
No. Mavick does not build or optimise schedules, and it does not forecast. Your master plan stays the plan, wherever you keep it. The agent answers from the copy you upload or connect: free line-days, allocations, plan against actual. It writes a re-plan proposal with the playbook’s options and every assumption stated, and the planner chooses. Any change it drafts runs only if an admin approves it. If you plan in scheduling software, export the plan to Excel or CSV and upload it, or keep it in a Google Sheet. Mavick does not connect to planning software or planning boards.
Our master plan is a multi-tab Excel workbook. Can the agent use it?
Yes, within limits. Upload it, up to 10 MB, and every sheet is searchable text. Only the first sheet is read as a table for sums and counts, up to 5,000 rows, so put the tab you calculate from first, or export it as CSV. An uploaded file is a copy: upload it again when the plan changes. If the plan is a Google Sheet, connect it, and the agent reads the cells when you ask and cites them. Either way, each answer names the file or sheet it read, so keep one current plan in the department, or put the version in the file name.
Can it move a production order in our ERP after a re-plan?
In Odoo, yes: it can draft a new planned start for a production order, and who it is assigned to. A department admin approves or declines it, it runs only if approved, and nobody edits a waiting action. It does not create orders, change quantities or move work between looms or lines in the ERP. A new delivery date for the buyer is merchandising’s call. Only Odoo production orders can be changed this way. With another ERP, the agent drafts the new start as an email to the unit head instead, sent only if approved.
Our hourly output is on a whiteboard. Can the agent still help?
Only once the counts are typed into a sheet. The agent cannot read a whiteboard, a paper plan card or a photo of either: a photo has no text to read. When supervisors’ counts go into a Google Sheet, or an Excel or CSV file you upload, it answers from those rows, and from an uploaded file it works out plan adherence. It does not connect to machines or counters, and it reads no loom data from the ERP: export loom efficiency to Excel or CSV and upload it, or keep it in a Google Sheet.
Will it warn us when a line falls 15% behind plan?
No. It answers when someone asks, and it does not check the plan between questions. The playbook’s 15% replan signal is a rule your planners apply. Upload today’s plan-versus-actual sheet and ask “which lines are more than 15% behind plan?”: the agent reads each line’s planned and actual rows, gives each line’s gap and names the lines more than 15% behind, citing the rows. From a connected Google Sheet, it reads the plan and actual cells, or a variance column if you keep one, and cites them. If today’s actuals are not there yet, it says so.
Can the planning agent see sourcing’s yarn dates or merchandising’s buyer forecasts?
Not their files. Each department’s data stays inside it, by design: the planning agent reads Planning’s own files, sheets and connected tools, documents shared with the whole factory, and Odoo production orders and stock transfers. A yarn receipt booked in Odoo is a stock transfer, so the agent gives its scheduled date and source document, with the fetch time. The supplier’s commitment behind that date belongs to Sourcing & Procurement, and buyer forecasts and order changes to Merchandising. To plan around a date or forecast it cannot read, put it in your planning sheet or the question, and it is treated as a stated assumption.
Bring Planning & PPC. We’ll build it on the call.
Book a 30-minute demo (opens in a new tab)- support@mavick.si
- Demo calendar
- calendly.com/safock/30min (opens in a new tab)