Research Article
Monitoring Big Data Streams Using Data Stream Management Systems: Industrial Needs, Challenges, and Improvements
Table 2
A summary of the challenges that an industrial company may face when implementing the required monitoring systems.
| | Challenges | Industries | Technical development |
| | Large dimensionality of monitored variables | x | | | High sampling rates | | x | | Searching data streams in a resource-constrained environment | x | x | | Handling huge amounts of equipment | x | x | | Real-time monitoring | | x | | Expert knowledge is limited-fault patterns not predefined | x | | | The dynamics of a system cannot be modeled (i.e., complex system) | | x | | Fault patterns expensive to simulate | x | | | False alarms | x | x | | Addressing concept drift | | x | | Model parameter estimation | | x | | Failures may occur suddenly (in short time) | x | x |
|
|