凯捷（Capgemini）：2023年产业研发数字化发展提速报告（英文版）（18页）.pdf

1、DIGITAL ACCELERATION IN INDUSTRIAL R&D How to rapidly and safely deliver value from data science and AI projectsCONTENTSINTRODUCTION1.PROVE VALUE BEFORE YOU COMMIT 2.IMMEDIATELY ACCESSING THE RIGHT DATA 3473.THE RIGHT TYPE OF INTELLIGENCE114.DEPLOYING MODELS AT SCALE14BRINGING IT ALL TOGETHER FOR RA

2、PID RESULTS16INTRODUCTIONDigital R&D is advancing rapidly,creating lucrative opportunities for innovation,by speeding up the time to get new innovations to market and reducing wasted effort and cost.Data science and AI are the jewels in the crown of Digital R&D,making it possible to spot unseen rese

3、arch opportunities,be more market driven,predict success or failure early,automate arduous processes,find new insights and evidence in small data sets,and optimize product development.Getting this right is not just about what is technically possible.It is about being able to use these tools in ways

4、that deliver tangible value to R&D,in timeframes that make it worthwhile.Data science and AI are complex tools that must be carefully integrated across different areas of R&D,and carefully aligned to the context in which they operate.Thought must be given to the whole implementation process from dat

5、a gathering,to model selection,to user experience.As Digital R&D accelerates,data science and AI will have an ever-greater role in doing R&D at speed,but must not compromise accuracy.Meanwhile,expectations of these tools will become increasingly complex,as the low hanging fruit of process automation

6、 gives way to more complex and nuanced uses of AI to predict product formulation outcomes.Data science and AI will be called upon to do ever more complex tasks,with ever less certain data.Organizations will find themselves with a constantly evolving portfolio of data science projects,which move them