In the experiences of Aimirim company through the years building and commissioning advanced controls, optimizations and digital twins for industries, we witness several gaps involving connectivity and integration between the plant operational technologies and information technologies mainly when the subject is to retrieve data from industrial systems and store them in a common place to enable data analysis and insights. While the IT world has a lot of companies investing in open source development with robust applications built on top of them with years of stable operation and continuous improvement softwares in industrial plants are becoming outdated and useless. It appears to us that those systems cared more about locking everything to themselves other than enabling interoperability to provide application of new technologies that the industry 4.0 needs.
Usually software in operational technology world keep some locks in:
- Communication protocols: Specific protocols developed by verndos do not provide interoperability. Only read/writes between the same vendors.
- Database: Vendor specific data structures limit use of data visualization, scientific computing and machine learning. Only software provided by the same vendor is capable of showing the data to you. And it does in an old, non-intuitive and ugly tool.
- Flexibility: Exporting data is too hard, when it even exists, becoming an exhaustive duty.
- Connectivity: Vendor's database is completely sealed. It does not offer a single form of accessing it from outside.
- Storage: The common practice in this area is to offer per-tag billing, even when the server is a stand-alone and belongs to the industry.
- Installation: Configuring a system like this usually requires a rare knowledge of Windows DCOM systems.
- Updates: If you want to access newer features from the vendors you need to replace all your PLCs with the latest vendor version and buy the license for the softwares all over again.
- Monolithic architecture: Hard to scale and low reliability when large amounts of data is needed.
- Deployment: Does not use container technologies to simplify deployment and updates.
Seeing that we came to a simple question that the industrial vendors forgot to answer: “To whom does that data belong?”.
For us at Aimirim it belongs to the industry itself, and therefore they should be able to gather, save, and access it as pleased.
This is no new problem, for quite some years the IT community has been focusing on observability, thus several open-source projects now exist to tackle each part of the problem.
Eclipse Tocandira will gather those tools and help you configure a scalable and open source historian to collect data from different PLC vendors, save it as much as the industry hardware allows to, and grant the user all the connectivity it deserves.
Eclipse Tocandira is a modern stack of open source software, based on microservice architecture to improve the observability focused on the industrial world. The aim of the project is to provide a way to retrieve data from industrial systems (PLCs, PACs, Gateways, OPC Servers) and store it in a modern time series database that could be in the cloud or on premise systems. All configurations are simplified by a meaningful UI with good user experience.
Eclipse Tocandira is a collection of tools aiming to help industries to remove their barriers on observability. To achieve this goal it uses the cutting edge technologies in the field. You will find the deploy easily configured with Docker, the connectivity with PLCs powered by 4diac, the gathering and storage done by Prometheus and the visualization shown with Grafana.Ec
Eclipse Tocandira was build focusing on:
- Reusability: Every technology chosen for this stack is the cutting edge of its field and the majority of them are well known in the IT world.
- Microservice Based: All services composing this application are truly independent and linked only by configurations, APIs and protocols.
- Cloud Native Architecture: These packages are usually applied for observability in IT world, and are therefore fully compatible with cloud environments, from instances to databases
- Small servers: Its components can run though several machines and consume little resources. It even runs on single board computers.
- Unlimited storage: Enough of charging for every variable read, the limit is only your hardware or the cloud capacity.
We think that the industrial world is too tightened to its vendors thus making it look like the vendors actually own the industrial data, and that needs to change.
It is needed to spread the world that a modern observability stack can help the industry too and without coupling it to a complete update of the hardwares involved. With the help of Eclipse Foundation we could promote a change in this market, popularizing Eclipse Tocandira might be the push that these traditional vendors need to change their behavior and start contributing to the open source community as well.
There are no legal issues to our knowledge
After the approval of this project we plan on have until the half of 2023:
- Open at least 2 of our connection blocks to different PLC vendors as contributions to the Eclipse 4diac Forte communication layer.
- Developments on the reliability of the stack integration to keep things running smoothly on connection failures.
And until the end of 2023:
- A Fully functional and friendly user interface to configure this whole stack.
- The first build of the whole system.
After letting this core as open source we plan on building a whole set of tools that will lead this to evolve in the community:
- Complete user and development documentation.
- CI/CD pipelines.
- Test suite.