Distributed Computing in Kreisel Energy Solutions

Kreisel Edge Nodes using Azure IoT Edge: a case study

Abstract
Software becomes the driving business factor in almost all industries and currently most proprietary systems hold on for dear life. Our goal at Kreisel is to develop an open software ecosystem that interconnects hardware devices of arbitrary type and class through a cloud platform with an application layer by moving a portion of the code base towards the "edge". We do that so we can leverage two major benefits:
Firstly, we effectively remove hardware dependencies. Devices, to which we have only protocol-level access ("black boxes") rather than code-level access, can be connected easily to the ecosystem by deploying a protocol translator next to it.
Secondly, we can reduce TCO (total cost of ownership) as well as backend software complexity by reducing data rates, data volumes, message frequency, VM compute time, VM compute demands, storage capacity, etc. That is achieved by moving even more functionality "to the edge", such as data preprocessing, data aggregation and machine learning.
For us, edge nodes in most cases are standalone devices like a RaspberryPi, an Arduino or any other industrial board with an OS and they are typically mounted in relatively close proximity to the actual hardware device. In addition to creating what's essentially a connectivity solution for our own products we take it one step further and want to open up our ecosystem to any kind of device and provide connectors for any software service on top.