A Practical R&D Expenditure Statistic and Management Method Based on Spatial-Temporal Representation of Multi-factors and Data Twin Technology

Nowadays, R&D Expenditure plays an important role in more and more creative activities of enterprises and other entities, especially in research activities and programs of society. However there still have a big problem that is how to collect and classify R&D Expenditure accurately. In this paper, after analyzing the restrictive collection factors on R&D Expenditure statistically, a practical scheme was provided that including R&D Expenditure Feature Vector and “Object Wood” concept were defined firstly, intelligent receipt recognizing model (IRPM), intelligent receipt persona model (IRRM) based on spatial-temporal representation of multi-factors and R&D expenditure data Twin(REDT) based on data multi relationship were developed creatively. Besides, intelligence carrier-class R&D expenditure management system (REMS) was developed based on above novel technologies and deployed on cloud with SaaS mode. For calling advantageously and updating conveniently, API standard interface and Full Stack Security Mechanism were also improved and used in REMS. Meanwhile, it was proved that REMS had better performance on assisting enterprise in collecting and using their R&D Expenditure after REMS employed by 50 industrial enterprises at first batch in practical over a period of time. There also have better economic benefits and social benefits after REMS was used by 211 enterprises in practically. Next, REMS would be utilized and tested in more scope of important entities so that the correlation technologies could be tested, iterated and optimized forward in the future. Actually, REMS is becoming R&D Expenditure industrial promoted by investor and market. Eventually, REMS would become one of the best R&D Expenditure collecting and using tools, it would not only promote R&D Expenditure increase but also become a industrial correlating with R&D Expenditure.