How can AI Benefit Product Development?
Manufacturers are deeply interested in how AI can benefit their product development. They are asking crucial questions: How can AI help us move faster and deliver productivity gains? What does the future of our business look like with AI systems automating operations? How can we leverage AI today for measures that impact the bottom line—cost, quality, and efficiency?
At PTC®, we have a clear point of view. The north star guiding this view is the Intelligent Product Lifecycle (IPL), where product data flows reliably across teams and systems from product engineering to after-market services. The IPL connects information like requirements, design intent, configuration plans, or service insights into a trusted data backbone. AI fuels innovation by embedding insights and orchestration across the lifecycle, taking shape as AI-powered features that advise, assist, and automate work.
Proven AI use cases are already delivering value today, from industrial copilots to engineering assistants for coding or requirement authoring. We’ve previously shared How AI creates value for manufacturers and compiled the latest market research, highlight how AI will deliver value across domains.
But how does this vision take shape? Where do you get started? What are the benefits and enabling technologies? These are the questions we will answer in this article, centering on an AI maturity framework that illustrates how AI-powered solutions impact manufacturers today and will deliver even more value in the future.
Introducing the AI Impact Horizons Framework
To help organizations understand how AI will transform product development, PTC has created the AI Impact Horizons framework. This framework structures AI-based impact into three time-based horizons:
1. Horizon 1:
Turbocharge Individuals: This initial stage is delivering value today. AI advisors and assistants boost productivity by increasing insights and executing specific tasks, such as validating an engineer’s requirements or optimizing a technician’s delivery schedule.
Underpinning these use cases are technologies such as natural language chat interfaces, LLMs, and AI assistants. Overtime tech adoption will include specialized models and automated agent workflows, boosting worker productivity with higher quality and speed.
2. Horizon 2:
Elevate Enterprise Intelligence: The next stage of innovation will deliver value over the next 1 to 3 years. It centers on connecting operations and breaking down data silos, enabling faster decision-making, and deeper orchestration across lines of business. Use cases such as enterprise change management or supplier design collaboration are already taking shape as early market proof points.
With evolving technology such as multi-agent coordination and cross product knowledge graphs, these AI use cases will take even more root in the enterprise, enabling faster decision-making and deeper orchestration across lines of business.
3. Horizon 3:
Reinvent Product Lifecycles: This highly mature horizon is predicted to generate benefits at scale in the 2-to-5-year timeframe. Integrated, autonomous AI systems will work together to reshape how product development takes place.
Built on advanced technologies such as agent networks, enterprise data meshes, and the convergence of digital and physical systems, industry will redefine product lifecycles with vibe product engineering and physical AI use cases.
As AI advancements are moving incredibly fast, these timelines from which manufacturers will realize benefits are rules of thumb. But what is clear are the use case patterns, organizational benefits, and technology enablers. In the following sections, we will dive into each horizon to identify these key elements, and along the way we’ll show examples from our work at PTC.
Horizon 1: Turbocharge Workers
The first horizon of AI transformation is defined by how intelligence empowers the individual worker. Across product development, AI is beginning to reshape daily tasks by providing instant access to knowledge, reducing friction with software adoption, and executing repetitive manual work.
Use Case Patterns
The most visible pattern is knowledge advisors. Using natural language, workers can ask questions, troubleshoot, find information instantly, or summarize files. For example, in PTC Windchill engineers can ask questions relevant to compliance documents or search for a torque value based on a product specification.
Ex: Use AI to ask questions and summarize documents in Windchill
A second pattern is workflow assistants, where AI executes specific tasks in a business process. Assistants can help with software coding, part classification, 3D model analysis, or service order recommendations. They reduce manual steps, perform validations, and guide workers through best practices. For instance, in PTC Codebeamer, requirements engineers benefit from several AI assistants that evaluate requirements for quality and help generate test cases.
Ex: AI Assistant in Codebeamer orchestrating test case generation
Over time, these assistants will evolve into more capable agents that can execute sequences of tasks—preparing a change summary, generating an ECO draft, running simulations, or compiling a service report. Ultimately, some workflows will become automated end-to-end, with humans approving and validating results rather than constructing every step manually.
This is the maturity model within Horizon 1: from knowledge advisors to task assistants to workflow automators. This progression applies to PTC products. A simple way to structure this is the AAA framework, which describes AI’s increasing capabilities starting with Advise, then Assist, and finally Automate—a topic we wrote about in this 2024 article, and organized in the table below by PTC product area.
Key Benefits
The benefits of Horizon 1 are immediate and significant. Workers save time by spending fewer hours hunting for data or navigating complex tools. They click less, switch context less, and complete tasks faster. Software becomes easier to use, making advanced capabilities more accessible to junior engineers or new technicians. Project timelines accelerate because bottlenecks—e.g. requirements cleanup, part searches, drawing corrections, service diagnosis—shrink significantly. Quality improves as AI catches inconsistencies, flags risks, and ensures more complete information at the point of decision. In aggregate, organizations see higher throughput, fewer errors, faster onboarding, and more consistent use of the systems they’ve invested in.
Technology Enablers
Several foundational AI technologies enable Horizon 1 use cases.
1. Foundation Models:
Models like GPT and Claude provide general-purpose intelligence, giving companies faster access to AI capabilities at a lower cost without building complex model infrastructure. Previously, building AI required deep expertise, proprietary data, and significant compute infrastructure.
2. Natural Language Processing (NLP):
Software requires users to know where information resides and how to navigate menus. NLP changes this dramatically. Workers can ask questions conversationally and invoke actions through plain language to retrieve results easily, effectively becoming power users of enterprise software.
3. Agents:
Agents put our enterprise data to work. They are designed to perceive their operating environment—e.g. prompts, context window—plan and execute actions, and improve future use through feedback loops—e.g. Reflexion. These agents bridge gaps between tools, augment expertise, and reduce time spent on administrative or coordination-heavy work.
4. Semantic Embeddings:
This technology paired with vector databases makes unstructured content—images, text, video—accessible to AI. This shift transforms knowledge access, enabling new use cases such as part search from 2D images, text-driven similarity retrieval, and richer document analysis.
5. Specialized Models:
Foundational models lack the domain context needed for product development. Specialized models close this gap. For example, vision models detect similarity part geometry or fine-tuned LLMs learn domain heuristics such as automotive design rules. They improve precision, reduce hallucinations, and in some cases lower cost through private hosting options.
Horizon 1 demonstrates how AI delivers immediate, practical value by scaling access to knowledge and reducing friction in daily work. These early gains are not just productivity wins; associated technology investments in data and AI set the foundations required for broader impact. As AI moves beyond individual tasks, the next horizon shifts focus from personal efficiency to enterprise-wide intelligence.
Horizon 2: Elevate Enterprise Intelligence
Horizon 2 represents the shift from personal to organizational productivity, where AI connects data and workflows across business departments. This unlocks more value from the digital thread beyond just data access, turning it into an intelligence layer operating across the product lifecycle able to deliver insights and automation.
Use Case Patterns
The first pattern is enterprise insights, where AI acts as a digital thread advisor. AI can retrieve and synthesize critical information across systems like PLM, ALM, SLM, QMS, ERP, MES, CRM, and FSM, answering cross-domain questions such as:
- “Which requirements are impacted by this design change?”
- “Can my suppliers meet these material specifications?
- “What are common service failures for this part?”
Traditionally, accessing these insights required report creation at each system, collating and curing data, and in some cases simply isn’t practical. AI accelerates access to information while operating across data silos—turning fragmented, effort-intensive investigations into shared insights that align interdependent teams.
For example, at PTC we built Asset 360® where field service planners and technicians can ask questions and build reports on the fly using an AI canvas. It combines natural-language-chat interfaces, with generative report building, all powered from FSM, CRM, ERP and PLM systems. Enabling technicians to quickly understand part requirements for a maintenance procedure, or service managers back in the office to understand spare parts inventory with intuitive charts.
Ex: with Asset 360 compare asset distribution based on product data
The second pattern is enterprise orchestration, where AI coordinates and executes multi-step actions within a cross-functional business process. A natural evolution of the task assistance and automation within a single product from Horizon 1—but now enabled across business units. Example implementation patterns include:
- Cross-system validation: AI collects and benchmarks data from two systems—such as comparing test results associated to product requirements
- Workflow progression: AI advances task completion—such as stewarding a change request by gathering impact data from PLM and cost inputs from ERP
- Decision automation: AI accelerates decision making by identifying missing information, proposing routing options, or recommending resolutions
An example of enterprise orchestration from PTC is AI-powered change management in Windchill. PLM and ERP agents automatically interpret a change request, identify impacted objects, summarize downstream effects, and coordinate approvals. Tasks that typically required repetitive, manual effort become streamlined processes, accelerating change steps and lowering coordination overhead.
PLM & ERP agents work together to accelerate change processes in Windchill
Another new example from PTC is centered on supplier collaboration. Here AI in PTC Creo and Windchill will streamline reviews by summarizing design inputs, automate work approvals, and recommend best-fit suppliers based on past projects and part requirements.
Key Benefits
Horizon 2 drives transformation at the organizational level by removing process delays and hidden costs caused by siloed processes. It elevates how the enterprise functions, turning fragmented operations into a coordinated, intelligent system. Time to market accelerates and costs drop while product quality and compliance improve. AI continuously validates dependencies across systems, aligning teams around shared context, and connecting test results and service feedback back into engineering—finally making the long-promised closed-loop lifecycle a reality.
Technology Enablers
The enablers for this horizon focus on agentic orchestration, integration, and data accessibility. Note these enablers can be applied in Horizon 1, but they aren’t required. However, at this stage, these enablers are essential to deliver enterprise intelligence:
1. Agentic AI:
AI agents collaborate across end-to-end processes, acting as intelligent data messengers and task implementers between systems. Specialized agents perform domain tasks, while orchestrators sequence work, verify outputs, and involve humans at key checkpoints. Governance layers enforce security, traceability, and compliance, enabling scalable automation with human oversight.
2. Modern Integration Protocols:
Protocols like MCP and agent-to-agent patterns enable flexible, runtime interaction with enterprise systems. Rather than hard-coding logic into point-to-point APIs, agents dynamically discover and invoke capabilities—supporting customer-specific data needs while future-proofing integrations by minimizing technology lock-in. Protocols like MCP and agent-to-agent patterns enable flexible, runtime interaction with enterprise systems, future-proofing integrations.
3. Semantic Layer and Product Graph:
The semantic layer defines shared terminology and ontology of product data—what entities exist and how they are connected to one another. The product graph instantiates that meaning by populating the semantic layer with real product data, organized as data tables, enabling AI to answer complex questions and execute workflows grounded to real-world and up-to-date context.
Horizon 2 technologies turn fragmented lifecycle data into shared operational context, unlocking faster insights and coordinated decisions. With this foundation in place, Horizon 3 examines what happens when AI moves beyond orchestration to actively co-create and operate products across their lifecycle.
Horizon 3: Reinvent Product Lifecycles
Horizon 3 paints a picture of the future of product development, where AI moves beyond orchestration to actively co-create and operate products. It brings to life visions that have been imagined for years but are more attainable based on the incredible pace of AI innovation.
What is clear is pieces of this vision are possible today—viewing and annotating 3D representations of connected assets, threading data objects between systems using product graphs, AI that can generate production-ready code, or advanced agents that can self-improve. If we extend these patterns and apply them to the future of product development, two use case patterns emerge, each driven by a distinct form of intelligence.
Use Case Patterns
The first is Vibe Product Development (VPD), where users express intent—e.g. optimize cost, improve safety, generate five architecture options—and AI generates validated outputs in a semi- or fully automated method. This evolution of "vibe coding" focuses on creative intelligence.
A defining behavior will be iterative design–test loops, where AI generates concepts, tests constraints, validates compliance, and incorporates insights from simulation, manufacturing, and service data. In systems engineering, this aligns with a shift-left strategy of the V-model, where verification and validation traditionally performed downstream are executed earlier during product definition. Networks of agents validate intent through software tests and virtual CAD simulations, while enforcing configuration constraints from PLM and cost targets from ERP. By iterating digitally before physical build, teams reduce late-stage rework, risk, and cost.
At PTC we deliver engineering software that is foundational to vibe product development—ALM, CAD, and PLM. Moreover, we are embedding AI capabilities into our software, establishing the building blocks for advanced AI. For instance, in our CAD software we are on a journey to increase AI maturity starting with AI advisors that reactively support engineers (Horizon 1) to proactive assistants that integrate multiple inputs (Horizon 2), and ultimately design intelligence where CAD AI orchestrates the creation of 3D geometry—a key pillar for enabling VPD (Horizon 3.)
The second use case pattern is Physical AI, where physical assets—products, robots, machines, tools—can perceive their environment, reason about conditions, and act autonomously to improve their performance. Thematically, it centers on an operational intelligence.
This frontier AI differs from present-day robotics, which automate well-defined tasks in controlled environments. These systems are highly effective within their design bounds but lack an understanding of the broader physical context and struggle when conditions change. Physical AI, by contrast, aims to create machines that embed advanced AI models of the physical world, enabling reasoning about objects, forces, and other real-world contexts—and are specialized for their domains such as car fleet management, assembly systems in factories, or power systems in buildings.
Ex: Creo simulation capabilities are foundation technologies for Physical AI
This is where simulation platforms and Omniverse-style environments become essential. Physical AI systems learn and validate behaviors in high-fidelity virtual worlds before acting in the real one. Those worlds depend on authoritative digital foundations: precise CAD geometry, configuration and variant management from PLM, and a lifecycle state that reflects how products are built, deployed, and serviced.
PTC’s role sits squarely in this foundation. PLM provides the system of record for product structure, configuration, and change. CAD supplies the geometric and behavioral fidelity required for simulation and digital twins. Together, they form the content backbone that feeds physical simulations and enables safe, scalable Physical AI. This is why PTC is partnered with Nvidia on this frontier of AI.
Key Benefits
In Horizon 3, manufacturers dramatically change how products are created and operated. AI shifts creation from time-exhaustive, sequential engineering to intent-driven, AI-led design and test loops. Capable of drafting requirements, proposing architectures, generating 3D concepts, and running rapid simulations—letting teams explore more ideas in a fraction of the time.
At the same time, self-improving lifecycles introduce continuous optimization, where AI analyzes test results, supply chain events, service logs, and regulatory constraints to detect issues early, refine designs automatically, and push validated improvements upstream. Once products reach the field, Physical AI enables adaptive, self-tuning machines that learn from real-world data, optimize performance autonomously, and interact with operators through intelligent copilots. Collectively, these capabilities redefine industrial competitiveness through faster innovation, higher reliability, lower lifecycle cost, and dramatically smarter products operating in the physical world.
Technology Enablers
Given the pace of AI evolution, complexity in product development, and overarching uncertainty in markets today, the technology enablers for Horizon 3 are based on our best understanding of current trends and the technical underpinnings of each use case pattern.
1. World Foundation Models (WFMs):
These models understand physical behaviors, enabling AI to simulate and optimize outcomes before physical execution as well as operate machines in the field. Software vendors will deliver base WFMs, while enterprises fine-tune them with proprietary product and operational data.
2. Agent Networks:
Cooperative systems of agents collaborate through shared events and semantic context rather than hard-wired integrations. Network-wide governance—identity, policy, auditability, and safety controls—enable these systems to operate reliably at enterprise and ecosystem scale.
3. Omniverse and Simulation Platforms:
Horizon 3 relies on simulation environments where AI can explore, test, and validate decisions before acting in the physical world. For example, platforms such as NVIDIA Omniverse enable multi-physics simulation, synthetic data generation, and collaborative virtual environments—allowing AI to learn, optimize, and coordinate safely without real-world risk.
4. Digital Twins:
These establish consistent identity and behavior across the product lifecycle. Product twins preserve lineage across EBOM, MBOM, SBOM, and as-built and as-operated states, enabling traceability and compliance. Behavioral twins synchronize physics, logic, and telemetry, allowing AI to simulate, predict, and validate decisions—providing essential grounding for models, simulations, and agent coordination.
Conclusion: The AI Impact Horizons and the Opportunity Ahead
The AI Impact Horizons framework defines a clear path for how AI will reshape product development. It starts with individual productivity and quality improvements at the individual level, expands to enterprise-wide decision making and orchestration, and ultimately reimagines how products are created and operated. Each horizon marks a step change in impact—advancing from individuals, then organizations, and ultimately to entire value chains.
What’s clear is manufacturers who embrace this shift will have a competitive advantage. And it’s not optional. The complexity and uncertainty of modern product development will only intensify, and the organization’s that leverage AI to overcome these challenges will sustain and expand the value they deliver to the market.
To capture this opportunity, leaders can act now by focusing on four strategic moves:
1. Enrich Your AI Strategy:
Use the AI Impact Horizons framework to map your strengths—proprietary IP, domain expertise, high-fidelity data—to Horizon 1-3 opportunities. In parallel, identify capability gaps—data quality, fragmented systems, or insufficient AI foundations. A strong AI strategy clarifies where to invest, identifies key risks, and which long-term bets to enable.
2. Strengthen Your Product Data Foundations:
Accessible, structured, and governed product data underpins every horizon. Digitize core processes while migrating, cleaning, and enriching data. Advance data maturity across domains—part-centric PLM, linked requirements and tests, MBD-annotated CAD, or rich service records. These foundations give AI the context—the who, what, where, and when—required to deliver relevant and reliable results.
3. Deploy AI Advisors and Assistants:
Start where value is immediate. Expand individual capacity with AI knowledge advisors. Build on this with AI assistants that reduce rework and accelerate routine tasks. These early wins validate the approach, showcase ROI, and establish the AI tech foundations required to scale into Horizon 2 and eventually Horizon 3.
4. Future-Proof Your IT Backbone:
Prepare your IT architecture for the AI era. Strengthen AI governance with robust authentication and validation of AI systems. Modernize infrastructure by consolidating systems and embracing SaaS where possible to take advantage of faster innovation cycles. The goal is to build an IT backbone that can both safely and swiftly deliver value to the organization.
PTC is uniquely positioned to help companies navigate these horizons with confidence. With deep expertise across the product lifecycle, industry-leading software, and over a decade of applying AI—our proven capabilities accelerate adoption, reduce risk, and deliver measurable value. Whether you are starting with AI advisors or connecting enterprise workflows, PTC provides thought leadership and technologies needed to turn the Intelligent Product Lifecycle into a competitive advantage.
The opportunity is clear and the companies that act now will deliver the most value in this AI era.