From concept to market: How AI can accelerate physical product innovation
Intelligent agents are transforming the traditional product innovation life cycle across industries, shifting the opportunity from incremental improvement to radical disruption
Ben Dollar
Lynne Sterrett
Diana Kearns-Manolatos
Monika Mahto
Nirmal Kujur
Artificial intelligence is advancing rapidly, with generative AI models providing new opportunities for industry leaders to drive competitive differentiation through front-office technology transformation. Beyond AI enabling digital advancements, one significant opportunity for industry leaders is how they can use the technology to innovate in the physical world—either through breakthrough products or smarter operations that deliver a sustainable industry advantage.
Organizations are currently prioritizing the digital aspects of product innovation, with a notable share of annual digital transformation budgets now allocated toward advancing automation through AI.1 At the same time, gen AI’s advanced reasoning capabilities can help accelerate and even redefine the physical product development life cycle at every stage, from ideation and design to prototyping, mass customization, and distribution. This transformation goes well beyond what we’ve discussed in “From manufacturing to medicine: How digital twins can unlock new industry advantages.” Gen AI enables organizations to rapidly explore and validate design concepts, helping them more confidently make investment decisions before execution while accounting for real-world constraints. The technology has also helped organizations enhance their engineering change management, ensuring incremental innovations cascade smoothly into manufacturing, minimizing disruption, and optimizing capital allocation decisions.
Moreover, the rise of physical AI on the shop floor is helping drive innovation: Product development is evolving to redefine the very capabilities and flexibility of manufacturing operations. This evolution transforms the workflow into a triad of collaboration—engineers, operators, and AI systems jointly shaping manufacturing strategy in real time—enabling not just adaptation but also proactive shaping of capital investments and operational strategies.
AI technologies, including gen AI, agentic AI (which fuses autonomous action with intelligent reasoning), and smart robotics, are being adopted by leaders across industries. Gen AI is already widely used for automated quality analysis, enabling better defect detection, higher yields, and lower costs,2 helping reduce research and development timelines from years to months, and in sectors like pharmaceuticals, cutting average prototype development cycles by up to 70%.3 These advances hinge on three core gen AI capabilities—simulation, prediction, and optimization—that together enable faster, more accurate modeling, testing, and production of new prototypes.4 In materials testing, gen AI accurately models molecular structures and predicts interactions, reducing manual experimentation. Agentic AI autonomously designs and refines prototypes, while synthesis-pathway optimization, the sequence of chemical reactions needed to produce a target molecule or material, identifies efficient production routes, shortening the path from concept to prototype. While these benefits apply to both digital and physical product development, use cases related to physical products are currently relatively few and therefore discussed less often. Physical products are produced and operate in complex multi-physics environments that have to account for material properties and manufacturing constraints, and thus may be harder to capture in the text- and image-based data sets that AI models are typically trained on. Also, AI must integrate with established computer-aided design, computer-aided engineering, and product life cycle management systems to influence real-world design workflows. Fortunately, these integrations are gradually happening.
As enterprises experiment with gen AI use cases, one emerging application is the incorporation of image and text generation tools into the product development process. Organizations embracing these tools are seeing benefits such as faster iteration cycles, earlier detection of design issues, and significant gains in efficiency.5 Acting as a digital partner, gen AI enables teams to rapidly explore a broad spectrum of design options, uncovering novel concepts that might not have been considered otherwise. This enhanced approach to design and delivery allows engineers to detect and address potential issues early in development, reducing the risk of costly errors once products are deployed into production.6
- Design as part of concept generation: Colgate-Palmolive is leveraging gen AI across its innovation cycle, distilling consumer insights to identify unmet needs and swiftly generating new product concepts. The aim is to enable faster, market-driven innovation and sharper responsiveness to meet evolving consumer demands.7 Similarly, Insilico Medicine has applied gen AI to drug discovery, advancing phase II clinical trials with patients.8 This shift has reduced research and development timelines from years to months, potentially unlocking billions of dollars in value by accelerating market entry and improving patient outcomes.9
- Delivery of rapid prototyping and engineering: Johnson & Johnson has used gen AI to innovate the creation and viability testing of new drugs. The company uses data related to unique chemical biosignatures to drive intelligent reasoning engines in the search for the most viable molecules to speed up its bioengineering of new therapies.10 AI has helped accelerate and automate hypothesis generation and testing, shortening discovery timelines and improving candidate selection.11
- Deployment to improve product quality and after-sales delivery: GE Aerospace’s AI-enabled blade-inspection tool has been deployed on both narrowbody and GEnx widebody aircraft engines, halving blade inspection times while significantly improving accuracy compared with traditional borescope methods. These innovations get blades in action more quickly, supporting rising air travel demand and advancing industrial product performance.12
How does AI physical product innovation manifest across industries?
Analysis of Deloitte’s 2025 Tech Value Survey data indicates the extent to which respondents across different industries are open to experimenting with new technology in general and how they perceive the return on investment in AI, gen AI, agentic AI, and smart robotics.
How are AI and gen AI investments driving product value across industries?
Product innovation can be measured in many ways (such as sales of new digital products, R&D spend on digital technology, and digital product launch effectiveness). These dimensions were collectively examined in Deloitte’s 2025 Tech Value study to identify industry-specific patterns and uncover the factors that drive superior outcomes (figure 2).
What does measurable AI-driven physical product innovation look like in practice?
As AI drives breakthrough results across industries, the transformation of traditional product innovation processes manifests in different ways as leaders seek unique, strategic advantages.
- Life sciences and health care: According to Deloitte’s 2024 AI and Medtech Survey, the greatest value emerging from AI and gen AI is in product development (according to 42% of respondents). Medtech companies are revolutionizing concept design and prototyping for new physical devices. The survey suggests that AI could reduce R&D costs by up to 20%, which, for a large medtech company, amounts to potential savings of as much as US$300 million over the next two to three years.13 In health care, building consumer trust in gen AI is crucial to clinician and patient adoption,14 as organizations like the Mayo Clinic harness gen AI to drive innovation in drug discovery and precision medicine, enhancing disease understanding and personalizing treatments at scale.15 Beyond drug discovery, gen AI is also being used for product innovation in areas such as diagnostics. For instance, in 2024, academic medical center Northwestern Medicine created an in-house gen AI system to draft near-complete radiology reports, helping radiologists boost productivity by up to 40% without compromising accuracy.16
- Consumer: The Clorox Company uses gen AI to rapidly prototype product concepts and personalize marketing strategies. The company has invested in a gen AI innovation tool that identifies global trends, enabling it to quickly develop hundreds of digital prototypes simultaneously and test them with millions of consumers. By doing this, Clorox is able to gather consumer insights more quickly, refine new product ideas, and bring innovations to market faster, bringing to the fore the combined ingenuity of human marketers and AI-generated prototypes.17
Another example is in the automotive sector: Autonomous vehicles, which increasingly provide real-time data on geolocation and telemetry, are a prime candidate for future innovation across the value chain.18 In Detroit, the Michigan Mobility Collaborative launched Accessibili-D, an autonomous vehicle pilot program using AI-powered technology to connect senior citizens and those with disabilities to essential services. AI-enabled autonomous vehicles use advanced decision-making and real-time data analysis to navigate complex environments safely and efficiently. Part of this effort included a sophisticated data platform leveraging AI for real-time data integration and actionable insights, guiding continuous improvements in safety and rider experience. The pilot demonstrated how AI-driven mobility can bridge transportation gaps, garnering high citizen satisfaction and serving as a scalable model for other cities.19
- Energy, resources, and industrials: According to Deloitte’s 2025 report on the use of agentic AI in manufacturing, while only 6% of manufacturers currently use AI and gen AI systems, nearly a quarter expect to adopt these technologies within two years to streamline concept design, digital prototyping, production planning, and quality assurance in engineering physical products. Siemens is actively driving product innovation using generative and agentic AI, as demonstrated by its new Industrial AI agents. AI agents are also transforming manufacturing processes, smart metering, and equipment management.20 As an example of enhanced equipment management, one oil and gas company managing 1,000 pieces of critical equipment across 80 locations was able to use AI to take diagnostic information from an existing AI platform and summarize it using gen AI speech-to-text to improve equipment maintenance. This enabled human engineers to go beyond using an existing dashboard to summarize the most important issues and more rapidly diagnose and remediate physical component repairs for enhanced quality assurance.21
- Government and public services: Deloitte’s ongoing research on the impact of gen AI in the public sector illustrates how this technology can drive product innovation beyond commercial applications.22 In the context of urban infrastructure, potential applications could include policy creation, facility design, as well as urban planning. A leading example is Singapore’s Virtual Singapore initiative, which uses digital twin technology to create a comprehensive, real-time 3D model of the city, integrating data from sensors, satellite imagery, and geographic information systems to accurately represent both above- and below-ground infrastructure.23 Similarly, Dubai’s 2040 Urban Master Plan is leveraging gen AI to facilitate city design processes, enable city planners to validate AI-generated urban layouts, and encourage collaboration with residents, municipal officials, and visitors, helping shape the future infrastructure of the city.24
- Technology, media, and telecommunications: Live sports are a major entertainment driver within this sector, generating significant broadcasting and streaming revenue as technology and media distribution rapidly evolve. For example, in Formula 1, AI simulations are being used to build and test new car designs.25 In technology, semiconductor leaders are deploying gen AI not only to design advanced chips and optimize equipment performance but also to model and accelerate the construction of state-of-the-art fabrication plants (known as fabs), with human engineers overseeing decisions, according to a 2024 Deloitte report on gen AI transformation of the semiconductor value chain. Intel is leveraging gen AI to optimize fab construction, prevent delays, and enable predictive maintenance.26
Charting the path for AI + human ingenuity
AI isn’t just speeding up physical product development—it’s fundamentally reinventing how companies turn ideas into reality by adding automation and intelligence to the physical product design life cycle, stress-testing prototypes virtually, and optimizing manufacturing in real time, all before a single product hits the market. As gen AI models become multimodal (able to understand 3D geometry, materials, and manufacturing parameters), their use in physical product innovation is likely to expand rapidly. This new wave of AI integration is enabling organizations to launch novel products faster and with greater precision than ever imagined, transforming what innovation looks like in the physical world.
Yet, this leap forward won’t happen by accident: Trailblazing organizations are breaking down silos, supercharging their teams with continuous learning, and embracing rapid experimentation as the new normal. By weaving AI deeply into every step, they’re not just speeding up cycles or trimming costs but also uncovering smarter solutions, spotting flaws before they matter, and setting a new benchmark for reliability and breakthroughs in physical product innovation.
As humans and AI converge, what will define the next era of physical innovation—your team’s creativity, your AI’s capability, or the synergy between them?
by
Ben Dollar
Lynne Sterrett
Diana Kearns-Manolatos
Monika Mahto
Nirmal Kujur
The authors would like to extend special thanks to Jim Rowan and Ryan Kaiser for generously contributing their time and insights, enriching this work with valuable perspectives and depth.
The authors would also like to thank the industry leaders who contributed their time and insights, further strengthening this work.
We also extend special thanks to our Deloitte colleagues, Kimberly Barb and Wynne Robinson, for their steadfast support throughout the development of this article.
We would also sincerely thank our editorial team, Aditi Rao, and Prodyut Borah, for their collaboration and guidance.
Editorial (including production and copyediting): Aditi Rao, Prodyut Borah, Pubali Dey, and Anu Augustine
Design: Pooja Lnu, Molly Piersol, and Sylvia Chang
Cover image by: Jim Slatton; Adobe Stock
Knowledge Services: Rishitha Bichapogu
Visit the Deloitte Center for Integrated Research
Access more insights on some of the most complex issues facing businesses today.