Frequently Asked Questions

Q: What Analytics technologies are used (off-the-shelf vs. proprietary solutions)?

A: SORBA is composed by multiple technologies (proprietary technologies own by SORBA and open source technologies licensed under the Apache 2.0 or MIT). There is not any proprietary technology used in SORBA outside of the ones developed by SORBA. There are several open source technologies that has been modified to include new capabilities:.

Q: What are some example snippet of some of the open-source technologies leveraged in the SORBA-AI automation platform:

Spark
Node-Red
Cassandra
Hadoop HDFS
TensorFlow
Grafana

Q: What are example snippet of some of the proprietary technologies contained in the SORBOTICS-AI automation platform:

SORBA ML auto training
SORBA ML edge processing
SORBA ML Post-Processing
SORBA ML auto pipelines
SORBA Historian Big Data Schema
SORBA IoT Data Ingestion Process
SORBA SDC connectivity pipeline

Q: What are the validated cloud platforms supported and how is the cloud vendor independency managed? Especially around the device connectivity and data flow.

A: All the SORBA technology is 100% independent of the cloud vendors. SORBA can be hosted on any cloud provider such as Azure, AWS, Google and it can be hosted on premise as well. Device connectivity and data flow are independent as well of any proprietary technology or cloud providers.

Q: What kind of Edge Framework is used / What kind of asset management framework is used?

A: SORBA developed their own edge framework which contains multiple pipelines for data collection, auto ETL, data processing, visualization, etc. in addition to fully developing our own Asset Management Framework which is fully distributed, it provides the tools to create the structure of a full IoT and Analytics solution.

Q: Could you detail the level of process and solutions around data privacy and Cybersecurity?

A: SORBA aligns its network connectivity with the widely adopted ISA & DMZ Purdue Models. SORBA provides several layers of security and encryption, such as Oauth Client/Server authentication, SAML, OpenID Connect, SHA-256 data encryption, tunneling, UIDs and SSL.

Q: What technology is used for the ML “Smart agent”? Docker?

A: There are multiple IT technologies that are supported by the Smart Agents, example Docker, VM, Kubernetes, etc.

Q: Are the ML algorithms pre-defined or custom?

A: The SORBA ML Agents (SORBOTS) are composed by a proprietary pipeline that provides the flexibility of creating automatically the best possible algorithms pipelines to achieve the best results, these pipelines can be customized, or they can be pre-package and reuse if it is needed (Transfer Learning).

Q: Can the smart agent be deployed outside the SORBA Platform?

A: Yes, the algorithms can be use outside of the SORBA Platform using Python as the recommended language, although other languages can be used if wrappers are created. 

Q: Can we add custom connectors? Connectors for IT Systems such as MES, CMMS? Can SORBA push data to other cloud or on-premise databases such as i.e. Azure IoT Hub?

A: Yes, this is all possible through the workflow orchestration layer that enables full 3rd party business system integration capabilities.

Questions and Answers