智能化创新：人工智能如何变革研发

How AI is transformingResearch and Development Table of contents Executive Summary31Behind the curtain: Today’s reality in R&D and operations4Industry challenges5The AI revolution in R&D62AI to support within the product lifecycle7Key valuable use cases within the product lifecycle, available today8Building an AI-enabled R&D organisation143Successful AI implementations are business-led, not tech-led174Conclusion20 Executive Summary As industries face increasing pressure to innovate swiftly and sustainably, integrating artificialintelligence (AI) into Research and Development (R&D) processes has become essential. AI has the By transforming vast data into actionable insights, AI enables organisations to streamlinedevelopment, enhance resource efficiency, and bolster compliance. Key applications that areachievable with today’s state of technology include variant management, requirements engineering, The PwC framework lays the groundwork for overcoming implementation barriers and fostering aculture of continuous innovation, positioning organisations for long-term success in an increasingly Behind the curtainToday’s reality in Industry challenges As the 2025 study from PwC in collaboration with MicrosoftAI inoperations: Revolutionising the manufacturing industryhasshown, the discrete manufacturing industry is under immense Time to market has become a decisive factor in competitiveness.PwC's research from an upcoming study on the future of R&Dshows that companies must accelerate their development cycles toseize opportunities while balancing speed with quality and Compounding these pressures is the scarcity of skilled talent.There is high demand for engineers, designers, and specialists inadvanced manufacturing, which limits the capacity to scale Together, these factors create a challenging environment in whichincremental improvements are no longer sufficient. Companies At the same time, AI presents a unique opportunity. Our surveyreport,AI in operations: Revolutionising the manufacturingindustry, produced together with Microsoft, shows that artificialintelligence fosters innovation in business environments by The AI revolution in R&D R&D is the cornerstone of innovation and at the heart of thesechallenges, making it an ideal area in which to apply artificialintelligence. AI has the power to transform the way organisations By analysing large and complex datasets, AI enables faster conceptvalidation especially for complex systems, predictive modellingand optimised design processes, reducing reliance on costly The sustainability benefits are equally compelling. AI can supportthe creation of resource-efficient designs and the optimisation ofmaterials, as well as the development of circular product Integrating AI into R&D is a strategic shift, not merely atechnology upgrade. It strengthens innovation capabilities,shortens development cycles and enables manufacturers to AI to support withinthe product lifecycle Key valuable use cases withinthe product lifecycle, available today In today's dynamic business world, the use of AI is becomingincreasingly important to enhance the efficiency andcompetitiveness of companies. Particularly in the context of the Each of the five main phases of the product lifecycle—innovation,product development, realisation, order process, and phase-out—presents challenges that are solvable with today's state of AItechnology. These technologies empower companies to streamlineoperations, anticipate market needs, and swiftly adapt to change. AI to optimise variant andcomplexity management In product development, mastering variant management is oftenchallenging due to its inherent complexity, necessitatinginnovative solutions in R&D. The key lies in effectively managingexternal and internal complexities—balancing portfolio, module, Effective variant management distinguishes ‘high runners’ from‘low runners’ and links these to technical implications, such as thenumber and severity of component variants required. Misjudging Current AI solutions tailored for R&D, like PwC’s METUS*, addressthese challenges by optimising product configurations usingadvanced algorithms. They help reduce unnecessary complexity,minimise component counts, and maintain a wide range ofconfiguration options. METUS leverages AI to quantify the impact AI to drive requirements engineering Requirements engineering provides the foundation for product development. Weak requirementsengineering can slow progress and increase risk. It can lead to fragmented communication, Operationally, poor requirements management drives delays, costoverruns, and resource inefficiencies. Misalignment among teamsand a lack of clarity on project objectives further exacerbate these Generative AI addresses these issues by enabling instant extractionand consolidation of requirements using semantics and opticalcharacter recognition (OCR) to translate handwritten informationinto machine-readable text. Our work with clients demons