Copy of Platform

Created by Raúl Galán
Jan 30, 2023
4 min read

Monom is an Industrial DataOps platform (data, apps and AI) focused on providing specialist adaptable solutions for predictive maintenance, quality and process enhancement.​​

The ability to merge and process data from different systems, combined with a digital twin model which can accommodate both data and features such as AI or physical models, makes it possible to obtain a holistic vision of the past, current and future behaviour of assets and their environment.

Global Architecture

The platform is divided into several parts according to the path that the data follows from its acquisition to its processing, storage and consultation by a user.

Regarding to “sources”, MonoM can ingest both IT data (e.g. SQL, API, SFTP, S3...) and OT data (industrial protocols, ModBus and OPC UA, and access to sensors and devices, thermometers, accelerometers, cameras...), as well as OT data (industrial protocols, ModBus and OPC UA, and access to sensors and devices, thermometers, accelerometers, cameras...)

Edge has several functions: transport information from the devices to the cloud, algorithms execution of machine learning models or devices managament (for example, the edge can send a signal to a PLC)

When data arrives to cloud, first we find four blocks:

- Machine learning service, training and execution of ML

- Dataflow, that is responsible for connecting the data ingested from the physical world to develop a digital twin

- Catalogs, that allows third-party systems to access information via API (assets, alarms, data...)

- Data lakes, where data is stored

The last part consists of displaying the processed information through:

- Dasboards, configurable dashboards with different types of graphs allowing data exploration

- Reports, allows for the creation of reports based on Google Data Studio / Google BigQuery, offering users the possibility of preparing almost any kind of report.​

- Alarms, creation of alarms based on conditions

- Multichannel alerts and notifications

Sources

IT Sources

Monom offers the possibility of acquiring data from multiple sources, provided that they use industry standards such as REST API, SQL, S3 or others. This makes it possible to combine OT data with other types of data, thus allowing users to send all of their important information to a single point. ​ ​

Generally speaking, these types of integration will require developments by Monom, as it is possible that a process will be required to adapt IT data to the format expected by the platform’s API. Finally, specialized connectors are available for some industries, such as connectors capable of reading and interpreting the data provided by the CMS of the main wind turbine manufacturers.

Monom can ingest data from several sources and known protocols. Data acquisition can be done from an Edge or from the Cloud itself and sometimes a connector may be required.

OT Sources

Monom has connectors that allow for its inter-connection with various industrial devices.​

The way in which these connectors are deployed is based on Microsoft’s Azure IoT Edge technology, which offers the possibility of managing connectivity from a single point.​ Monom provides to their client an Azure IoT Image ready to work, and may supply edge computer as well if it is needed

At present, Monom is capable of managing the OPC UA and Modbus industrial protocols, along with some from other manufacturers (DATOM, Sensorworks) 

 

 

 

Monomedge  

Monom uses Azure IoT Edge for its edge management, which ​allows it to perform the following tasks​:

  • Device management​

  • Deployment of connectors and adapters​

  • Data ingestion ​

  • The possibility of executing analyses of different kinds at the edge itself, thus improving response times.

​Finally, our edge allows for events to be received from the platform in order to trigger configurations or actions on the edge or on plant components.​

​It allows for the installation of hardware of various kinds, for example computers with processors of different architectures, depending on the needs and user requirements of each project.

 

Data Services

Analytics Services

The product offering is divided into 4 groups depending on the type of user:

  • Online learning, Maintenance Managers / Plant Managers

  • AutoML, suitable for any profile of the organization

  • DIY, Data Scientists

  • Box Algorithms, suitable for any organizational profile

Dataflow

 

Visual tool that provides users with the ability to transform the data acquired from the sensor into information that is ready for exploitation using flow composition. One of the most important functionalities consists of the possibility of executing scripts, offering users the power to transform the data according to their needs. It also allows variables to be combined with the aim of obtaining new variables derived from the originals.

Other feature is the possibility to execute Machine Learning Models.