Sneak Peek · AIVA APS

Know what's short, why,
and what to do next.

An AI-native planning engine that answers the three questions every manufacturer asks: which orders ship on time, what material is short and why, and what to buy and when — with full traceability to prove it. Currently in development — here's a first look at what's coming.

AIVA APS · Schedule Results
98.2%
On-time delivery
1,284
Orders planned
14
Shortages detected
23
Late orders
Late Orders
TOYOTA · TR-X-2200 9 days
DENSO · DN-118 6 days
AISIN · AS-77 1 day
Shortages
A-1234 Material short
C-771 Lead time
H-30 Supplier delay
Procurement Actions
A-1234 · Expedite Urgent
C-771 · New PO Planned
H-30 · Confirm ETA Follow up
What AIVA APS does

Full-factory MRP/APS — one engine, every answer.

AIVA APS takes your factory's real snapshot — demand, inventory, purchase orders, WIP, BOM, lead times, calendars — and runs a complete material requirements planning engine. No spreadsheets. No guesswork.

Demand Planning

Normalizes and prioritizes every demand source — MPS master schedules, sales orders, and forecasts — into a single ranked queue with configurable priority rules.

Supply Allocation

Pools on-hand inventory, in-transit shipments, open POs, and existing manufacturing orders — then allocates to demand by priority, availability date, and material attributes.

Multi-Level BOM

Processes multi-level bills of material top to bottom — handling phantom items, engineering changes, alternative materials, and intermediate inventory offsets at every level.

Intelligent Scheduling

Schedules backward from due dates and forward from real material availability — then picks the honest planned start based on what's actually possible, not what you wish were true.

Shortage Detection

Identifies exactly which materials are short, how much is missing, which orders are impacted, and whether the shortage is structural (lead time exceeds the window) or solvable.

Procurement Recommendations

Generates actionable purchase suggestions — which item, from which supplier, when to release, when it arrives — respecting MOQ, lot sizes, lead times, and receiving dock calendars.

Demand Input
MPS · Sales Orders · Forecasts
TOYOTA · TR-X-2200 · 800 pcs MPS
DENSO · DN-118 · 1,200 pcs SO
AISIN · AS-77 · 450 pcs FCST
HINO · HN-44 · 220 pcs SO
Supply Pool
Inventory · POs · Manufacturing Orders
A-1234 · 3,500 on hand Stock
B-892 · ETA 06/15 PO
C-771 · In production MO
D-445 · In transit Transit
How it works

End-to-end planning in five stages.

AIVA APS runs a complete planning engine against your factory snapshot. Each stage feeds the next — no manual handoffs, no intermediate exports.

1

Snapshot

Freeze demand, supply, BOM, routings, calendars & parameters into a reproducible planning run.

2

Rank & Allocate

Prioritize demand, match planning attributes, search primary & alternative supply by time bucket.

3

Explode & Schedule

Explode multi-level BOMs, schedule backward from due dates and forward from real material availability to find honest planned dates.

4

Plan & Peg

Create planned production & procurement orders. Build full demand-to-supply traceability across all BOM levels.

5

Validate & Publish

Automated validation — demand coverage, quantity checks, supply realism, delivery performance — then publish results.

Physically honest. AIVA APS never hides lateness. If supply can't arrive before the due date, the plan says so — giving you time to act instead of a false sense of security. Every timing calculation respects factory calendars, receiving dock days, and supplier lead times from your master data.

Full traceability

Every allocation has a paper trail.

Traceability is a first-class output, not an afterthought. AIVA APS records the complete supply-demand relationship across every BOM level — so you can trace any shortage back to its root cause and any procurement recommendation back to the demand that triggered it.

Top-down

Start from any sales order and trace down: which production orders serve it, what components they need, where each material comes from — inventory, PO, planned procurement, or shortage.

Bottom-up

Start from any supply — a delayed PO, a depleted stock lot — and trace up to every demand it impacts. Know immediately which customers feel it.

Evidence-based

Every allocation link records the type, quantity, timing, rule applied, priority rank, and reason. Quantities are validated automatically: nothing gets lost or double-counted.

Demand-Supply Pegging · Trace View
Sales Order
TOYOTA · TR-X-2200
Due 06/20 · 800 pcs
LATE 3 DAYS
Planned Production
MO-4471 · Assy TR-X-2200
Start 06/12 → Complete 06/23
800 PCS
▼ BOM Explode
Component · A-1234
Housing Frame
Need 2,400 pcs by 06/12
ON HAND 2,400
Component · B-892
Control Board
Need 800 pcs by 06/12
PO · ETA 06/15
Component · C-771
Sensor Module
Need 800 pcs by 06/12
SHORT 200 PCS
Procurement Action
C-771 · Planned PO
200 pcs · Supplier: Murata · LT 12 days
RELEASE BY 06/04
Fulfilled Late supply Shortage Planned action

Why pegging matters: When a supplier says "your PO is delayed 5 days," pegging tells you in seconds which sales orders are impacted, which production orders are waiting, and whether the delay cascades into customer-facing lateness — or is absorbed by buffer time. No manual tracing. No guesswork.

What you get

Answers, not raw data.

AIVA APS produces structured decision-support outputs — each one answering a specific question your planning team asks every day.

OutputAnswers
Master Schedule Every demand's completion date vs. due date — on time or late, and by how much.
Short Orders Which materials are short, how much is missing, and which demand caused the shortage.
Procurement Plan What to buy, from which supplier, when to release the PO, and when material arrives — with late-arrival alerts when arrival exceeds need date.
Production Timing Planned start vs. latest start for every manufacturing order — highlighting orders that can't start on time due to material constraints.
Traceability Map The full demand-to-supply causation tree — multi-level, multi-source, with ECN substitution paths and quantity validation at every link.
Delivery Performance On-time delivery rates by scope (SO, forecast, all), week-rolling trend, with V4 physical-honesty scoring.
What-if simulation & AI planning agent

Ask the plan. Get grounded answers.

AIVA APS includes an AI planning agent that answers questions grounded in your actual plan data and traceability — not generic model knowledge. And a what-if workbench that lets planners test scenarios before committing.

Ask "Why is this order late?"

The agent traces the supply chain, identifies the constraining material, shows the supplier lead time vs. the available window, and explains whether the delay is structural or solvable.

Test "What if this PO is delayed?"

Positive actions (expedite, air-ship, substitute material) run in seconds. Disruptive changes (rush orders, supply reallocation) trigger a full re-plan with before/after comparison.

Compare plan versions

Every planning run is snapshotted and versioned. Compare two versions side-by-side — which orders improved, which degraded, what supply changed — with full evidence for each delta.

What-If · Simulation Result
23
Late orders (before)
19
Late orders (after)
Scenario: Expedite material A-1234
TOYOTA · 06/22 → 06/13 On time ✓
DENSO · 06/10 → 06/12 Late 2d
AISIN · 06/14 → 06/16 Late 2d
Impact summary
Orders rescued 4
New lateness introduced 2
Net improvement +2

Questions the agent answers: Why is this sales order late? Which supply fulfills this demand? Which PO delay impacts which orders? What planned procurement was created and why? What changed between two plan versions? — Every answer cites specific data from your plan, not general knowledge.

Built for manufacturing reality

Enterprise-grade planning rules, not toy demos.

AIVA APS implements a comprehensive set of business rules refined against real factory data — covering edge cases that spreadsheet-based planning and generic MRP systems miss.

Multi-source demand

MPS master schedules, sales orders, and forecasts each follow distinct due-date conventions. AIVA APS normalizes them into one ranked priority queue while preserving source traceability.

ECN & substitution

Engineering change notices with effectivity dates, alternative material rules, and old/new part convergence — so shortage detection doesn't create false positives from part-number transitions.

Pool-based allocation

Supply is pooled and allocated by priority — not locked to individual orders. MTO surplus flows back to the shared pool after bound demand is satisfied, maximizing material utilization.

Calendar-aware timing

Factory calendars, shift patterns, receiving dock working days, and supplier lead times are all read from master data — never hardcoded. Reruns are reproducible against the same snapshot.

Procurement intelligence

No POs in the past. Same-week same-item consolidation. MOQ and lot-size rules. Release-date clamping. Arrival-vs-need late alerts. Supplier-specific lead times from vendor master.

Four-layer validation

Every demand accounted for. Quantities balanced across the plan. Real vs. planned supply clearly separated. Delivery performance tracked weekly. No unchecked gaps.

Roadmap

MRP core in development. Finite capacity scheduling next.

The material planning core is under active development. The engine is designed in layers — each one building on the last toward a full enterprise APS solution.

⚡ Layer 1 — MRP Core In Development

Demand planning, supply allocation, multi-level BOM processing, full traceability, intelligent scheduling, planned production & procurement, shortage detection.

Layer 2 — Constraint Propagation

Feasible time windows for production operations. Forward and backward propagation across operation precedence to eliminate impossible schedules before optimization.

Layer 3 — Finite Capacity

Work center load profiles, overload detection, bottleneck identification, resource-constrained scheduling, and Gantt-ready operation timelines.

Ready to see what your plan really looks like?

AIVA APS is available for early-access manufacturers. Talk to us about running it against your factory data.

← Back to AIVA