Various hardware modules are incorporated in SpinSplit’s flow chemistry ecosystem such as pumps, thermostats, valves, electrospinner units. Each of such devices are controlled by edge computing units, which are implementing the physical level communication with the actual hardware actuators and sensors. Once a task is executed at the end-point hardware, its operation should be always supervised at the edge level through the available sensor devices.
The goal of this individual project is to design a container-based data storage system for the supervision subsystem of the edge computing units of SpinSplit LLC, Hungary. The storage system is a non-relational database by requirement, which can deal with heterogeneous data.
The project consisted of:
- Choosing a suitable noSQL database system and justifying the choice by conducting performance analysis and suitability testing for real-time operation of the Edge Supervisor.
- Define the system requirements and build the block diagram of the integrated Edge supervisor database to the existing edge computing system.
- Designing algorithm for the database operations based on the system and operational data collected from the flow chemistry instruments and edge computing units.
- Implementing the designed algorithms for the database in python in an OOP style and compile a new python library called “eSupDB”.
- Perform post-implementation performance analysis of the written code and the database.
- Participate in the deployment process of the database in the SpinSplit testing environment.
