The invention discloses a method for directly accessing hardware data into a
big data platform. The method comprises the following steps: acquiring hardware equipment data and sending the data to an
Internet of Things computing gateway; performing
edge computing on the acquired data by
The Internet of Things computing gateway, compressing the high-frequency repeated data according to change conditions, and performing computing to obtain prediction and alarm data; Data after edge calculation and prediction and alarm data are submitted to a
big data warehouse of a distributed architecture, and achieving real-time warehousing of the collected data.
The Internet of Things computing gateway based on the
ARM architecture has the advantages that the overall architecture of
the Internet of Thingscomputing gateway based on the
ARM architecture for compressing and computing the front-end acquired data reduces the dependence on a
cloud server, reduces the expenditure of the
server, and is quickin response and high in reliability. According to the method, multi-dimensional information prediction based on
artificial intelligence and deployment of the distributed bins are realized in a front-end severe environment, and the problems that a
computer based on an X86 architecture cannot meet the front-end severe environment and a computing
queue is too long, prediction cannot be performed in time and the bins cannot be distributed discretely due to
centralized computing of a
server are solved.