Nomura: China Factory Automation Faces 2-3 Year Painful AI Transition; Bullish on Inovance's Hardware Integration Path

Overview

Nomura Securities released a research update on China's factory automation and industrial AI implementation. The core view is that despite the immense potential of industrial AI, its commercialization does not occur as a standalone product but relies on upgrades to the legacy Distributed Control System (DCS) installed base. Domestic Operational Technology (OT) incumbents, leveraging vertical industry expertise, are superior to general-purpose IT platforms in AI deployment. However, due to structural constraints such as security concerns from State-Owned Enterprise (SOE) clients, resistance to domestic substitution, and cross-vendor protocol incompatibility, the industry must endure a 2-3 year 'painful transition period' before reaching an AI-driven revenue inflection point. The report highlights the contrasting AI monetization paths of Inovance Technology (hardware physical integration) versus Supcon Technology (software bundled subscription) and maintains a Buy rating on Inovance Technology.

Core views

Core View 1: Industrial AI monetization is anchored to legacy DCS, not an independent growth engine. The report notes that industrial AI currently acts primarily as a Revenue Amplifier, monetized by bundling with the existing DCS installed base. For example, Supcon Technology packages its flagship AOP (Autonomous Operating Plant) with TPT (Time-series Pre-trained Transformer) and MaaS/RaaS subscription services for petrochemical clients. In Q1 2026, Supcon disclosed industrial AI revenue of CNY 184 million, accounting for 12% of total revenue; however, industry surveys suggest at least 20% of this is essentially bundled DCS hardware sales, indicating limited pure software subscription penetration. Core View 2: Inovance Technology and Supcon Technology exhibit distinctly different AI implementation paths. Inovance adopts a 'physical integration' strategy, embedding AI and computing power directly into servo drives, motors, and specialized PLCs in new energy and battery factories, entering the humanoid robot supply chain via integrated controller-servo solutions combined with real-time force control. The report believes Inovance monetizes AI through Hardware Pull-in rather than software subscriptions, placing it at a different AI tier compared to software-dominated process industry incumbents. In contrast, Supcon relies more heavily on software bundling and service subscriptions, deeply entrenched in the petrochemical sector. Core View 3: A 2-3 year painful transition period constrains the revenue inflection point. The report emphasizes that the Chinese automation market remains deeply tied to DCS, and AI has not yet become an independent driver. Two major structural hurdles exist: first, MaaS/RaaS subscription models face domestic substitution pressure and security compliance resistance from SOE clients, leading to some loss of traditional orders; second, cross-vendor compatibility is extremely poor—for instance, a Siemens PCS7 biopharmaceutical project stalled for six months because the Profinet protocol could not bridge with Supcon's Modbus, ultimately requiring a complete replacement. These factors imply the industry needs a 2-3 year adjustment period before achieving large-scale volume growth.

Analysis framework

The report employs an analytical framework of 'Tech Stack Breakdown + Business Model Comparison + Industry Survey Validation.' First, it defines the four software pillars of China's process industry AI stack (TPT, UCS, AOP, OMC), noting that domestic competition centers on vertical industry depth rather than horizontal model capabilities. Second, using recent industry survey data, it quantitatively analyzes the AI revenue composition of leading companies (e.g., Supcon), stripping out hardware bundling to assess true software penetration. Third, it compares the monetization logic differences between distinct technical routes (Inovance's hardware embedded vs. Supcon's software subscription) in specific downstream scenarios (New Energy vs. Petrochemical). Finally, combining specific cases (e.g., Siemens-Supcon protocol conflicts) and macro constraints (SOE security concerns), it deduces that the current revenue inflection point is delayed by 2-3 years.

Methodology notes

• Industry/Sector Analysis FrameworkSupply Chain Upstream-Midstream-Downstream Transmission

  Role differentiation between OT (Operational Technology) vendors and IT (Information Technology) platforms in industrial AI implementation

  The report points out that in industrial scenarios, OT vendors with deep vertical process know-how have an advantage over general IT platforms because AI implementation relies on understanding specific production processes. This reflects the logic of midstream manufacturing barriers transmitting to downstream applications.

• Company Fundamentals & Financial Framework

  Comparison of Hardware Pull-in vs. Software Subscription monetization models

  The report distinguishes two AI monetization paths: one drives hardware sales by enhancing hardware performance (e.g., Inovance embedding AI in servo motors), and the other relies on software licensing and service subscriptions (e.g., Supcon's MaaS). This classification helps understand gross margin structures and revenue sustainability across companies; software subscriptions typically yield higher marginal profits (≥20% vs. DCS <10%).

• Valuation MethodologyPE/PEG valuation

  Target price setting based on historical average PE premium

  The report uses relative valuation for Inovance Technology's target price, applying 39x 2026E EPS. This multiple is above its 33x historical average PE (+0.8 standard deviations), reflecting expectations for long-term growth and an AI premium.

Asset mapping & comparison

Structured mapping from thesis to named assets (strengths, weaknesses, peers, risks).

• Inovance Technology (300124 CH)

  Beneficiary; monetizes AI via hardware physical integration, entering humanoid robot supply chain

  Strengths

  Hardware DNA, AI embedded in servos/motors/PLCs, avoids pure software compatibility issues, deep roots in new energy sector

  Comparison

  Compared to Supcon, Inovance relies more on hardware pull-in than software subscriptions, operating at a different AI monetization tier

  Risks

  Market share loss, EV business gross margin decline

• Supcon Technology (688777 CH)

  Neutral/Watch; software bundling model constrained by SOE security concerns and protocol incompatibility

  Strengths

  Deep entrenchment in petrochemical sector, complete AOP/TPT software stack, large legacy DCS base

  Weaknesses

  Cross-vendor protocol incompatibility (e.g., Modbus vs. Profinet), MaaS subscriptions face SOE security resistance

  Comparison

  Software-led; significant portion of AI revenue still involves bundled hardware; pure software penetration needs improvement

Key data

• Supcon Q1 2026 Industrial AI RevenueCNY 184 Million12% of total revenue, of which ≥20% is bundled DCS hardware
• Inovance Target PriceCNY 85Based on 39x 2026F EPS, implying ~13% upside from current price
• Industrial AI Software Gross Margin≥20%Significantly higher than traditional DCS hardware gross margins of <10%, supporting blended margin expansion
• Transition Period Duration2-3 YearsExpected duration of the painful period before the AI-led revenue inflection point
• Inovance Current Share PriceCNY 75.08Closing price as of June 5, 2026

Impact & implications

For the industry, industrial AI will not bring explosive independent growth in the short term but will exist as a value-add to traditional automation equipment. Investors should lower expectations for rapid scaling of pure software AI revenue and focus on OT vendors with strong legacy DCS customer bases. For Inovance Technology, its 'Hardware + AI' physical integration path offers uniqueness in humanoid robotics and new energy sectors, potentially avoiding some compatibility and security audit risks faced by pure software subscriptions, thus maintaining relatively steady growth during the transition. For process industry vendors like Supcon Technology, the true subscription revenue inflection point awaits clarification on AI safety standards from SOE clients and unification of cross-vendor interconnection standards.

Risks

• China factory automation demand below expectations
• Inovance Technology faces market share loss from domestic and international competitors
• EV business gross margin declines more than expected due to intensified competition and rising raw material costs

What to watch

• Stickiness of industrial AI software subscription models
• Progress on cross-vendor protocol compatibility standards
• Policy changes regarding AI security compliance requirements for State-Owned Enterprises