Hoshin Kanri for the industrial CEO — the X-matrix with live KPI, not last month's.

The problem

The Hoshin Kanri cascade is perfect — until you have to measure it.

Hoshin Kanri (the X-matrix, the cascade of objectives from committee to plant and back) is one of the most solid Lean methods there is. The method is not the problem. The problem is that, the moment you have to measure whether the plan is on track, the data travels the way it did in 1995: someone at each plant consolidates a spreadsheet, someone else at corporate merges them, and the CEO looks at a chart two weeks after month-end close.

It is the same old syllogism: there is a strategic objective, there is an indicator at every workstation feeding it, there is data in SCADA/MES/ERP measuring that indicator. But between the data and the CEO's chart there are three people, two spreadsheets and a slide deck. That chain eats time and, worse, trust: when one plant reports differently from another, the entire Hoshin review is spent arguing where each number came from instead of discussing the plan.

And in multi-plant groups the problem multiplies: each plant consolidates by its own criteria, the corporate committee receives spreadsheets with different definitions, and the board debate turns into accounting archaeology instead of industrial strategy.

How it fits the IRIS system

iLEAN does not replace your Hoshin Kanri — it closes the cascade's measurement loop.

Hoshin Kanri is excellent at designing the plan and the cascade. What it lacked for decades is the central brain that keeps the KPI live without anyone consolidating by hand. That is the filler iLEAN provides — it seals the crack between the data already circulating on the floor and the indicator the committee needs, without asking you to change the X-matrix, the BI or the ERP.

The data lives in islands; the plan lives on a wall. iLEAN connects the islands and brings the plan live to the executive desk. The CEO discusses the plan, not the spreadsheet.

The three iLEAN pieces applied to real-time Hoshin Kanri:

• Connect — captures the data from every workstation and every plant wherever it lives: modern SCADA/MES, the panel of an old machine via photo, the controller's spreadsheet, a shift report dictated by voice. Standardized at second zero, without forcing you to change anything already in place.
• Brain — the central piece that holds up the X-matrix cascade. It maintains the map of objective → indicator → data source → accountable workstation, and aggregates in real time from the workstation to the chief executive. Same data, same criteria, every plant.
• Agent + Writer — the agent tracks deviations against the plan (and proposes actions to the plant manager when a KPI drifts); the Writer prepares the executive dossier for the committee and for the CEO, automatically, with live data and traceability behind every figure.

See the full IRIS architecture →

Before and after

Traditional Hoshin Kanri vs. Hoshin Kanri with live KPI

Aspect	Traditional Hoshin Kanri	With iLEAN Brain + Writer + Agent
KPI frequency to the CEO	Monthly, 1–2 weeks late	Live, same day
Cross-plant consolidation	Manual, different criteria per plant	Automatic, same criteria
Data origin	The controller's spreadsheet	SCADA/MES/ERP + capture of data islands
Hoshin review	Arguing where the data came from	Discussing the plan
Deviations from plan	Seen at month-end close	The agent flags them as they happen
Multi-plant sovereignty	Tension between plant and corporate	Security rings: each plant keeps its own

Impact estimate

Impact estimate for your plant — to be validated with your numbers.

The block below is an estimate to be validated with your group's actual data. We lay it out so the committee has an order of magnitude; we refine it during the diagnostic.

• Industrial group with 3–6 plants, annual X-matrix already in place, corporate BI in use, KPI hand-consolidated by the controller every month.
• Connect + Brain pilot on one plant and a bounded set of strategic X-matrix KPI (typically 5–10 KPI per level). First value expected within a few weeks: the CEO sees the first live KPI on their desk before the next month-end close.
• Reduction in consolidation time ahead of the Hoshin review estimated at ≥ 40%. Indicative payback between 4 and 9 months, depending on controller hours/month freed up plus the value of decisions that no longer wait for month-end close.
• The hard lever is not only saved hours — it is earlier decisions. A deviation from plan corrected the same day costs something very different from one corrected 30 days later.

And the external evidence that carries weight in the boardroom

The board's first question about an AI-fed KPI is always the same: can we trust it? The published research is clear that language models fail precisely when they are pushed to guess without a factual anchor ^[1]. That is why iLEAN never invents an indicator: every figure the CEO sees is anchored to a real, traceable source from the floor — SCADA, MES, ERP, a verified capture — and the chain from workstation to boardroom can be opened at any point. Real-time Hoshin Kanri is not one more system; it turns the data you already have into an earlier decision, which is what a board reads as sustained competitive advantage rather than IT spend.

^[1] OpenAI paper "Why Language Models Hallucinate", 2025 — on the reliability of AI in anchored tasks.