Source: sources/hitachi/2026-06-10-hitachi-dss-sector-briefing-abe.pdf
(25 slides; original filename 20260610_04_digital_en.pdf; translated from
the Japanese original, which prevails in case of discrepancy).
Presenter: Jun Abe, Executive Vice President and Executive Officer,
CEO of the Digital Systems & Services Sector, Hitachi, Ltd.
Supplied by Tell Münzing, 1 November 2026.
⚠ Conversion caution. Marker flattened most of this deck — it is
layout-bearing almost throughout, and the Markdown twin scrambles the figure
captions on pp. 2, 4, 6, 7, 9, 13, 14, 16 and 18 into adjacency without
structure. Nine slides were rendered natively with pypdfium2 and read as
images; every number and every quotation below comes from the rendered slide,
not from the twin. Do not cite the Markdown twin for this source without
checking the slide.
Why this matters: the altitude
This is the sector CEO, one level above Anand Birje and one below the group CEO. It is the only source the wiki holds that speaks for DSS as a whole. Where the April deck gives the group arithmetic and the October deck gives one BU’s operating picture, this one gives the thing both of them leave implicit: what DSS owns, what it sells to the other sectors, and what it is spending money on next year.
The numbers (p. 2)
| FY2024 | FY2025 | FY2026 plan | FY2027 target | |
|---|---|---|---|---|
| Revenue (¥bn) | 2,965.2 | 3,068.3 | 3,190.0 | CAGR 7–9% FY24–27 |
| Adj. EBITA margin | 14.2% | 15.5% | 15.7% | 16–18% (prev. 16%+) |
| ROIC | 10.3% | 11.5% | 10.5% | 11–13% |
| Lumada revenue ratio | 46% | 62% | 70% | 75–80% (prev. ca. 65%) |
All figures “based on the new reporting segment structure”; bracketed figures are the previous targets announced at Investor Day 2025 (11 June 2025). Backlog ¥1.8tn, +11% YoY. Group Lumada ratio 40%, +11pts.
⟨Inference: the Lumada target was raised by 10–15 points inside twelve months. Two readings are available and the source does not choose between them: either the Lumada business genuinely accelerated, or the denominator shrank when the segment was redrawn and the ATM business left. See open task 20 — this deck does not close it either, but it does show the revision happening in one step, which is more consistent with a definitional change than with organic growth.⟩
Drivers named for FY2025: contracts over ¥3bn in Japan at ¥330bn, +87% YoY; gross margin +1.4pts; defence revenue +26% (cybersecurity and intelligence, naval systems, vehicles); “top-tier profit margins among Japan IT service vendors”.
GlobalLogic: all of the growth is internal (p. 4)
The single most load-bearing slide in the deck for understanding the digital business.
| unit: $m | FY2024 | FY2025 | change |
|---|---|---|---|
| Stand-alone | 1,950 | 2,000 | +2.6% |
| Synergies (booked by other BUs & Group companies) | 1,450 | 2,500 | +73% |
| Total | 3,400 | 4,500 | +32% |
Footnote: “GlobalLogic contribution to the Lumada business. Synergies represent revenue recorded in each BUs & Group companies.” Caption: “The total is included in Lumada business revenue.” Stated challenge: “Transformation of stand-alone businesses.”
Three synergy case studies: a metaverse platform replicating nuclear power plants; a solution to optimize energy efficiency in train operations (Rail); a service using building facility data to optimize operations, maintenance and energy consumption.
⟨Inference, and it is the finding of this source: the external digital services business is flat. Every dollar of GlobalLogic’s growth came from selling into Hitachi. Birje’s “+46% YoY synergy channel against single-digit external” was the FY2026 half-year of the same fact, reported one level down; this is the full year, in the sector CEO’s own books, and it is more extreme. Any proposition to this organisation that arrives through the internal channel is arriving on the only growth path it has.⟩
⟨Inference, on Tell’s double-counting question: this slide is the clearest statement of the accounting anywhere in the sources. Synergy revenue is recorded in the receiving BU, not in GlobalLogic — so it is counted once in group revenue. What is counted twice is attribution: GlobalLogic claims it as contribution, the sector books it as Lumada. That is a narrative double-count, not a financial one, and it is exactly what the group-level footnote “reflection of revenue consolidation offsets started from this period” (April deck) exists to correct at the top.⟩
The market frame (p. 6)
“New business opportunities driven by AI: ¥100 trillion (2030)”, split three ways, all FY2025–2030 global, all from internal analysis of industry reports:
| Market | CAGR | Stated need |
|---|---|---|
| Modernization | 15–20% | “Reliable AI deployment in complex enterprise systems” |
| AI Services | 30–35% | (same column) |
| Social Infrastructure × AI → Physical AI | 15–20% | “AI deployment in mission-critical systems with ensured reliability and safety” — electric power & utilities, rail & transportation, manufacturing, government & public sector |
Hitachi’s four stated strengths beneath: large-scale high-complexity system development; global cutting-edge AI deployment and operations; capability to build and operate highly reliable social infrastructure; deep domain knowledge in OT and frontline operations.
⟨Inference: the fastest-growing of the three markets in Hitachi’s own analysis is AI Services at 30–35%, not the physical one at 15–20%. The Physical AI story is the differentiated one, but it is not the fast one. A proposition framed purely as infrastructure sustainability is being measured against the slowest of Abe’s three growth curves.⟩
The two frontiers (p. 7)
| FY2025 result | FY2025–2027 CAGR | |
|---|---|---|
| DSS sector AI-related revenue (modernization + AI services, global) | ¥800bn | 20–25% |
| Hitachi group HMAX revenue | ¥300bn | 50–60% |
Modernization/AI Services levers: “further deepen customer co-creation through the FDE team”; “transform into a continuous value delivery model (value-based and recurring)”. Physical AI levers: “deployment of Physical AI in social infrastructure”; “create value to innovate social infrastructure”.
⟨Inference: HMAX is group-wide and smaller than DSS’s own AI book by a factor of 2.7, but growing 2.5× faster. These are the two KPIs Abe is held to. A proposal that can be told as either is legible; one that is told as neither has no scoreboard to land on.⟩
Modernization: the installed base (p. 9)
≈15,000 customer systems in operation in Japan is the stated asset. Path: “enabling AI-readiness across extensive IT assets” → “AI deployment in modernized systems and delivery of AI-related services”. FY2030 Japan revenue target ¥3–5tn, with traditional SI and services flat beneath a growing Modernization and AI Services wedge (chart marked “for illustrative purpose only”). Key areas for strengthening: talent and the foundation for AI utilization.
FDE Team (pp. 10–11): AI specialists, domain experts, data scientists, security engineers, project managers, on-site with customers across the full lifecycle — problem definition → rapid prototyping → deployment and operation of mission-critical AI systems → continuous improvement. Target in Japan: 5,000 people, from a core workforce of 35,000 SEs in the DSS sector. Supported by the Frontier AI Deployment Center (announced 19 May 2026) providing technology, a use-case repository and on-the-job training.
Agentic Integration (p. 12): requirements-definition, design, coding, testing and operations agents running autonomous development against “Hitachi’s proprietary IP and assets”, with knowledge accumulation and asset reuse feeding back. Named assets: VelocityAI, HARC for AI, project management, quality assurance.
Physical AI: what DSS actually contributes (p. 13)
The same three-market slide as the April deck p. 20 — grid maintenance ¥15tn (2035), railway AI ¥3tn (2030), frontline worker support ¥10tn (2030) — but with the DSS contribution tagged explicitly on each:
| Market | HMAX product | DSS contributes | Hitachi’s physical base |
|---|---|---|---|
| Grid maintenance, power grid stabilization | HMAX Energy | Digital twin (DSS) | HVDC installed base over 150 GW; real-time grid control |
| Railway operational efficiency maximization | HMAX Mobility | Edge AI (DSS) | Signalling installed base over 30,000 km; autonomous and decentralized control |
| Sustained, autonomous productivity improvement | HMAX Industry | Autonomous evolution AI (DSS) | Customer Zero — 175 in-house manufacturing sites; replicating expert skills |
KPI: HMAX revenue growth 50–60% CAGR (FY2025–2027).
⟨This retires an inference. The wiki previously derived the DSS/sector division of labour by reading the April and October decks against each other — neither stated it. This slide states it: the sector owns the asset base and the customer, DSS supplies a named technology into each HMAX product. The operational rule stands and is now sourced: a proposal that needs a site goes to the physical sector; a proposal that is a mechanism goes to DSS.⟩
HMAX as an operating system — four pillars (p. 14)
“Evolve HMAX, which pursues ultimate reliability through four pillars, into the ‘OS’ that supports social infrastructure operations”, serving Energy, Mobility and Industry:
| Pillar | Named as | Definition on the slide |
|---|---|---|
| Domain Knowledge × AI | Physical AI | ”Integration of cutting-edge AI, on-site expertise, and OT” |
| Talent | Physical AI FDE Team | ”A team with deep expertise in IT, OT, and AI” |
| Security | Mission-Critical Security | ”Ensuring the safety of mission-critical infrastructure that must never stop” |
| Data Infrastructure | Data Fabric | ”Integrating and structuring IT/OT data, enabling the safe use of AI” |
⟨Inference, and the most useful single observation in this source: the sector CEO’s HMAX has a security pillar; the BU CEO’s stack does not. Birje’s four-layer diagram runs FDE service → AI Operations → Data Fabric → existing IT, with no security layer and no confidentiality, provenance or verification vocabulary anywhere. Abe names security as one of four pillars and funds “expand OT security services” (p. 18). The vocabulary SINE needs exists at sector altitude and disappears one level down. That is a routing fact, not a contradiction: the argument is legible to Abe’s organisation in a way it is not to Birje’s deck.⟩
⟨Inference, with care: “security” here means operational safety of infrastructure that must never stop — OT security, not data provenance or confidentiality between parties. The pillar is adjacent to SINE’s problem, not identical to it. Claiming the pillar would be wrong; attaching to the Data Fabric pillar and citing the security pillar as the reason is right.⟩
The Physical AI FDE Team is the DSS–sector interface (p. 15)
Drawn as two halves meeting:
| GlobalLogic + Hitachi Digital Services | → Physical AI FDE Team ← | OT Sectors |
|---|---|---|
| AI Technology · Security · Data Infrastructure | AI Specialist · Security Engineer · Data Scientist ↔ Domain Expert | Product Knowledge · Industry Domain Knowledge · Operational Technology |
Headline: “Advanced AI deployment in social infrastructure with enhanced reliability and safety”.
⟨Inference: this is the organisational joint between DSS and the physical sectors, and it is staffed rather than notional. It is also the only place in any source where a security engineer and a data infrastructure capability are shown physically sitting next to a rail or energy domain expert. For an approach that has to cross the DSS/sector boundary, this team is the crossing point — which matches the independent observation that an FDE team is already deployed inside Rail on Factory IQ.⟩
Data Fabric: the slide to show (p. 16)
Title: “Data Fabric Integrates IT/OT Data to Enable New Value Creation across Industries.” Section heading: “4.4 Data Foundation for the Deployment of Physical AI in Social Infrastructure”.
A red band labelled “DSS Sector” runs along the bottom carrying “Data Fabric (Data Infrastructure)”, with two bullets:
• Integration and structuring of IT/OT data • Enable cross-sector and cross-industry data utilization
Three arrows rise from it into a triangle of HMAX Energy, HMAX Mobility and HMAX Industry, with a named exchange on each edge:
| Edge | Value named |
|---|---|
| Energy ↔ Industry | ”Optimal production linked to green power supply and demand, etc.” |
| Energy ↔ Mobility | ”Dynamic optimization of railway power; localized energy production and consumption, etc.” |
| Industry ↔ Mobility | ”Full synchronization of production and logistics, maximizing just-in-time efficiency, etc.” |
⟨Inference, three steps and the third is the opening:
- Ownership is settled. The Data Fabric is DSS’s, by name, on the sector CEO’s slide. Not Energy’s, not Mobility’s, not CI’s.
- The purpose is already cross-boundary. “Cross-sector and cross-industry data utilization” is the stated job, and the first value example given is a green power / production coupling — a value-chain sustainability use case in Hitachi’s own hand, not one SINE has to introduce.
- But every arrow is Hitachi-to-Hitachi. All three edges connect HMAX products to HMAX products. There is no arrow to a supplier, a customer, an operator or a regulator. “Cross-industry” here means across Hitachi’s own industries. The architecture for data Hitachi is not given is not drawn — which is the gap the wiki has stated elsewhere as: Hitachi can integrate data it is given; for data it is not given, it has no architecture.⟩
Growth investment — there is a budget line (p. 18)
¥30bn organic investment (FY2026), plus “financial capacity for up to ¥1 trillion in inorganic investments (FY2026–FY2027)”.
| Talent | Technology |
|---|---|
| ”Strengthen and develop the specialists comprising our FDE team (AI, security, business analysis, project management, etc.)” | • “Enhance data fabric capabilities” • “Expand OT security services” • “Build an advanced AI utilization environment across the Hitachi Group” |
⟨Inference: “Enhance data fabric capabilities” is the first bullet of the technology column of a named, funded, in-year investment line owned by the sector CEO. This is the most concrete answer the sources give to “where would the money come from”. It does not mean provenance is in scope — it means the budget exists and its scope is being set now.⟩
Partners (p. 19)
Six, alphabetical: AWS, Claude, Google Cloud, Microsoft, NVIDIA, OpenAI. Headline: “Collaborate with partners to advance AI platforms and infrastructure, scaling value creation.”
⟨Inference: all six are compute or model-layer. No data-layer vendor is named at sector level — Databricks and Snowflake appear on Birje’s 15-partner BU slide but not on Abe’s six. The data layer is not vendor-committed at the altitude where the ¥30bn is allocated.⟩
Structural reform (p. 20)
Centred on IT Products. Storage: focus on market-leading block storage, ranked first in Japan’s enterprise external block storage market by vendor revenue (IDC, CY2025). ATM: “agreement on business integration with OKI, scheduled to start October 2026”. KPI: IT Products adj. EBITA margin +3.2pts FY2025–2026.
⟨Inference: the ATM exit is the concrete event behind the suspected DSS denominator change in open task 20. The deck confirms the transaction and its date but not its revenue effect.⟩
Risks, as stated (p. 22)
“Slowdown in IT investment” and “sharp rise in component prices”. Nothing about data access, regulation, trust or AI safety appears in the risk list.
What is absent
No circular economy, no sustainability, no emissions, no supply chain, no supplier, no provenance, no verification, no attestation, no confidentiality, no standards body. “Security” appears four times and means operational safety of infrastructure plus the OT security services line. The nearest adjacency in the entire deck is the phrase “optimal production linked to green power supply and demand” on p. 16 — one caption, no owner named.
Cross-references
- Hitachi, Ltd.
- Hitachi Digital — the BU inside this sector
- Anand Birje’s presentation, 28 Oct 2026 — one level down, four months later
- Inspire 2027 progress deck, 27 Apr 2026 — one level up, six weeks earlier
- Mobility
- Customer Zero
- Confidentiality gap in carbon accounting
- Open tasks — item 20
Sources
sources/hitachi/2026-06-10-hitachi-dss-sector-briefing-abe.pdf— the decksources/hitachi/2026-06-10-hitachi-dss-sector-briefing-abe/— Marker twin and extracted figures (see conversion caution above)