AI is moving rapidly beyond conversational tools and enterprise IT. Across manufacturing, organizations are applying it to product engineering, plant operations, asset performance and production processes, creating opportunities to improve development speed, product quality and operational performance.

Adoption alone does not guarantee value. HCLTech’s The AI Impact Imperatives, 2026 report found that when respondents were asked what percentage of major AI projects starting over the next 24 months they expect to ultimately fail, the average response was 43%. The challenge is connecting AI with the industrial data, engineering context, processes and business outcomes required to operate at scale.

This is why the next phase of manufacturing AI will increasingly be vertical-led, with solutions built around industry-specific engineering knowledge, data models, operational constraints and value drivers.

Vertical-led AI connects engineering context to business value

In manufacturing, AI-led engineering brings intelligence into the earliest stages of the product lifecycle, helping teams identify problems before they reach production, evaluate more design options and connect engineering decisions to real-world product performance.

The same principle extends across the manufacturing value chain. AI can support predictive maintenance, quality management, production optimization, asset performance and increasingly intelligent physical operations. The technology architecture may share common foundations, but the AI needs enough industry and engineering context to understand the processes it supports.

Our AI Force platform provides one part of that foundation. AI Force combines GenAI and Agentic AI to automate and augment workflows across software and data engineering, IT operations and enterprise business processes, with governance, contextual intelligence and autonomous agents built into the platform.

Knowledge graphs connect engineering context with AI

Industrial AI performs only as well as the context it can access. Manufacturing knowledge can sit across decades of computer-aided design models, product lifecycle management records, maintenance histories, operating procedures, quality records and field data, often fragmented across tools and generations of engineers.

Knowledge graphs can help connect that information by establishing relationships between requirements and components, assets and failure modes or design decisions and field performance. This gives GenAI and Agentic AI richer engineering context than isolated documents can provide.

For a manufacturer, that could mean connecting a design decision with historical quality issues, service records and field-failure data. The resulting AI system can help engineers understand why a problem occurred, identify related information and assess the potential implications of a new design decision.

This contextual foundation can become increasingly valuable over time because it reflects the manufacturer's own engineering knowledge and operating experience.

Improving decisions at the design stage

Many quality, cost, safety and manufacturing issues that emerge later in a product’s lifecycle can be traced back to decisions made during design and engineering.

Those decisions are becoming more complex. Engineers need to consider mechanical and electrical requirements, materials, weight, performance, sustainability, supply constraints and manufacturing processes. Understanding how these factors interact becomes difficult when the relevant information is distributed across different systems and teams.

AI can help correlate those inputs, identify patterns and surface trade-offs earlier. Generative design can explore multiple configurations based on defined requirements, while emerging text-to-design capabilities can translate engineering intent into initial options and AI agents can continuously validate designs against defined standards.

Engineering expertise remains central to this process. AI gives engineers a faster way to investigate possibilities, test assumptions and focus human judgment on the decisions where it creates the greatest value.

The result is a more connected design-to-manufacturing process where potential defects, cost pressures and production constraints can be considered before a product reaches the factory floor, when changes are generally easier and less expensive to make.

Connecting digital twins, AI and model-based engineering

Greater value emerges when AI is integrated with digital twins and model-based engineering.

Digital twins have long supported simulation and product validation. AI expands their potential by helping organizations analyze greater volumes of engineering, manufacturing and operational data within a connected digital representation of the wider system.

Generative design can create and refine possible solutions, model-based engineering maintains relationships between requirements, components and system behavior and digital twins simulate how those decisions may affect product performance, manufacturing operations and downstream service requirements.

Together, these capabilities allow manufacturers to test more scenarios before changing physical products or production environments. Engineering teams can examine how a design modification might affect weight, performance or manufacturability, while factory teams can model the consequences of changing equipment, production sequences or material flows.

This convergence is becoming increasingly important to the industry's direction. In its Manufacturing Predicts 2026 webinar, Gartner says that by 2030 manufacturing will be transformed by semi-autonomous AI agents, software-defined products and closed-loop digital twins, while rising IT costs and governance challenges will also need to be managed.

Why successful pilots struggle to scale

Many manufacturers began their AI journeys through proofs of concept led primarily from a technology perspective. These initiatives could demonstrate that a model performed a particular task, but that did not necessarily establish whether it could improve a business process consistently and economically.

Scaling requires a different starting point. Organizations need to define the business problem first, whether that means reducing engineering cycle time, improving first-time quality, preventing equipment failure or accelerating the introduction of a new product.

The model, data and technology architecture should then be selected around that objective. Without this connection, an AI initiative may work technically while remaining isolated from the workflows, ownership structures and decision-making processes required to create measurable impact.

At HCLTech, we focus on AI that delivers ROI: prioritizing practical, scalable use cases according to business value and establishing the foundations required to make successful solutions repeatable. This aligns with the terminology and positioning defined in our AI/GenAI brand guidance.

Our broader capabilities can support different parts of that environment. AI Force provides GenAI and Agentic AI capabilities across enterprise workflows. xLMCloud supports cloud transformation for engineering and product lifecycle management environments, while our M360 Manufacturing Suite provides end-to-end visibility across smart manufacturing operations.

The objective is to create reusable foundations that allow teams to move from isolated experimentation toward enterprise capabilities.

Bringing Physical AI into the engineering loop

Physical AI extends this transformation from digital analysis into physical environments.

Robotic systems, autonomous platforms and increasingly capable humanoid technologies can capture information and perform defined tasks in environments that may be difficult, repetitive or unsafe for people. Potential applications range from equipment inspection and environmental sensing to material movement, logistics and production support.

These systems can also create a richer feedback loop between engineering and operations. Data collected from physical assets and production environments can update digital models, helping engineering teams understand how products and processes behave under real-world conditions.

Effective collaboration between people and AI-enabled systems will be central to this evolution. Physical AI can perform defined tasks, gather information and respond to changing conditions, while employees provide contextual judgment, oversight and intervention where required.

Building the foundations for scale

AI-led engineering depends on well-structured industrial data. Manufacturers often hold valuable information across engineering systems, product lifecycle management platforms, manufacturing execution systems, enterprise resource planning applications, quality records, service histories and connected assets.

Possessing the data is only the starting point. Organizations need to connect it, preserve its engineering context and make relationships traceable across the product lifecycle. Models also need to be validated, monitored and updated as products, operating conditions and source data change.

Governance is equally important. Organizations need clear boundaries defining which decisions AI can support, which actions it may perform autonomously and where human authorization remains mandatory. Cybersecurity, model performance, traceability and accountability need to be built into the engineering architecture from the beginning.

Measuring engineering impact

Leaders should evaluate vertical AI-led engineering through operational and business measures rather than the number of models or pilots deployed.

Development speed can be assessed through engineering cycle time, simulation coverage, prototype iterations and the time required to evaluate design changes. Quality measures can include defects identified before production, rework, scrap and issues reaching the field.

Cost measures should consider engineering effort, physical prototyping, production disruption, warranty exposure and the ongoing cost of operating AI models. Product performance can be assessed through reliability, yield, energy efficiency, sustainability and service outcomes.

Competitive advantage will increasingly depend on how effectively manufacturers connect AI with engineering knowledge, industrial data and measurable business priorities. By combining digital twins, model-based engineering, generative design and Physical AI within a governed architecture, manufacturers can move from experimentation toward more intelligent, scalable and adaptive product development and operations.