Manobiak Engine

bunch-of-Apple-devices

Powered by 

credit-logos-small2

The Manobiak Engine is a cloud based data analysis platform and our research in progress. The development of the platform is motivated by the desire to simplify and automate workflows related to data storage, data analysis, signal processing, and machine learning. The platform is accessed directly via web browser on any type of device, meaning no software installation or licens management. 

Key features
  • Cloud native (HW-independent)
  • No SW installation needed
  • Out of the box data storage with backup
  • Share data, analysis setup and results with colleagues across the globe
  • The pipeline concept (Run data processing in parallel without bounds)
  • Customizable, scalable, and flexible
  • Build on open source libraries
  • State of the art Machine Learning algorithms

Watch a demostration of the modal analysis branch of Manobiak Engine and find more details about Manobiak Engine below.    

Manobiak Engine

Modal Analysis
KEY FEATURES
  • State of the art system identification algorithms
  • Automated analyses
  • Data storage upload / download
  • Complete traceability of data storage and analyses
  • Simple and user friendly interface 

More details

Manobiak Engine is a Python Flask web application currently running in the cloud. We use the platform as a service (PAAS) Google App Engine. Our software can also be deployed as a Docker container.

We are running on FireBase NoSQL, and on cloud stored HDF5 files (NCSA file format for large amounts of data). Orchestration of Google cloud functions is done using the Google PubSub service (MQTT-like broker service) for maximum scalability.

Visualization is done using d3.js and Matplotlib. All system identification is performed using an inhouse developed system identification library containing both publicly known and proprietary algorithms.

We always strive to become vendor independent to save money and to avoid breaking changes. We continuously try out new techniques, and technologies. Currently we are trying out running on InfluxDB and interfacing to AWS/S3. We are also testing various learning based extensions to our automated identification processing pipeline.

Manobiak ApS
Lergravsvej 2
DK-8660 Skanderborg
CVR-no.: 4173 3748

Contact information
info@manobiak.com