A featured contribution from Leadership Perspectives: a curated forum reserved for leaders nominated by our subscribers and vetted by the Construction Tech Review Advisory Board.
Beijing PERCENT Technology Group
The five trends in the application of data intelligence technology in the next 3-5 years
Liu Yijing
Companies around the world are speeding up the digital and intelligent transformation, focusing more and more on business intelligence.
Make CIO go to the front desk, find out the defects in business process and operation in time, and solve the problems through enterprise digital top-level design and IT technology. We believe that in the next 3-5 years, the application of data intelligence technology presents five trends:
Firstly, the fusion of data, technology and scenario.
In the process of digitalization and intelligentization, it is necessary to optimize business process, analyze business problems, understand business trends, assist business decision-making and finally bring action through the new generation of information technology, which is a complex closed loop of scenario. The complexity of the application results in the complexity of the data requirement, for it is necessary to combine the business data, log data(machine data), IoT data(sensor data), tagged data(human experience), simulated data and knowledge data to achieve the goal of digital transformation and intelligent transformation.
Thus, this trend is application-driven. The fusion is embodied in the fusion of multi-technology, multi-data and multi-application scenario. There is a need to integrate upstream and downstream technologies, including IoT, edge computing, cloud, speed protection, Big Data, business intelligence, artificial intelligence and AR/VR/MR etc.
Secondly, increasingly emphasize on real-time of application.
With the acceleration of digital transformation, more and more application scenarios need to respond in time, even in real time. It is urgent to change from T+1 to T+0, which requires high infrastructure and data technology.
Real-time is the trend of data intelligent application in the future. With the increasing demand of real-time and the increasing amount of data processing, the real-time-computing-related technology will become more and more popular. Spark, Streaming and Flink will be used more and more widely in the future, and may even replace Hadoop and MapReduce.
The third trend is comprehensive interactive penetration of AI.
Self-help customer service has gradually replaced people's repetitive work, we can get quickly the desired answer in the machine answer way. When data intelligence technology is gradually applied in various fields, non-professionals can also use data intelligence to assist decision-making.
Interaction is more and more natural, flexible and efficient between man and machine, between man and organization and between organization and organization, which requires technology to creatively overcome sensory barriers such as time, space, language, vision, hearing, touch and smell.
The fourth is machine automation and autonomy.
In the past, data intelligence technology can only perform precisely defined tasks. Now, disinfection robot, dispensing robot and feeding robot are more and more applied, they have, in certain degrees, the autonomy in specific task field and routine.
In fact, AI in essence is to constantly summarize the rules, with more and more data precipitation, the machine presents the trend of autonomous evolution, help people to make more choices and judgments. Autonomy is reflected in the system's ability to carry out some tasks in the closed loop of perception, cognition, decision and action after proper authorization.
The fifth is the edge computing empowering data privacy protection.
Data security and privacy are issues that need to be addressed in the next 3-5 years. In the past, the traditional way is to gather the data to the cloud for centralized processing, analysis, modeling and application. With the enhancement of edge node and device capability, the enterprise will run different prediction analysis and AI model on the edge, and can perform more operations on the terminal or at the edge node without uploading the privacy data.
