Data Management in Edge Computing

Why data management?

Data serves as the engine of digitization. Only on the basis of data, states of systems can be analyzed, visualized and controlled in a suitable manner.

By data management we mean the

  • Collection
  • Transfer
  • Saving
  • Linking
  • Enrichment with more information and
  • Visualization

of data.

Our know-how puts us in a position to extract relevant data from any system on-site, to link it with each other and, among other things, to realize novel services on this basis.

Connectivity and interoperability

In order to get the right data, it is necessary to be familiar with the various communication interfaces. These can be used to call up data from sensors, operate actuators and establish communication between different systems. A particular challenge here is to establish the interoperability between different communication interfaces. The retrofitting of older existing systems often requires the fast connection of proprietary interfaces that do not meet today's communication standards. A high priority is also placed on the security of collecting, storing and passing on the information, so that no unauthorized person can gain access to this information or manipulate it.

Our focus is on communication interfaces that are used in the following application areas:

  • Industry and building automation
  • Energy systems
  • Internet of Things (IoT)
  • Connection to cloud solutions
  • Internet and web services

Handling with metadata

In addition to the acquired analytic data, additional information is needed for learning data analytics algorithms as well as evaluating the information. These so-called metadata must also be recorded and stored with the user data in a suitable manner.

For example, metadata can contain the following information:

  • Description of the system state (production system is in regular operation or shows a specific error)
  • Place of data collection
  • Time of data collection
  • Responsible for the data collection

Visualization of data

Visual inspection of data prior to data analysis often provides important starting points for choosing the right analysis algorithm, but also for operating a system. It must be distinguished whether a user interface with a corresponding user experience is to be realized or the data is to be examined offline in a computer for special features.