When working with data from industrial assets, whether from automation systems or directly from the micro-controllers, it is crucial to be able to create structures in your back-end that allow you to describe data from a business context, and not just a 1:1 rendering of “temp_sensor_123” from “machine1234abc”. In other words, you want to be able to flexibly define structures that suit the individual employee, customer or business partner. It is also often the case that the there are multiple organisations involved in the use case scenario and they even have different internal roles and thus different requirements of how they wish to use data. It is therefore important that an IoT platform provides flexibility in terms of structuring data in many different ways and that the underlying data model is easy to integrate.
Let’s explore this further:
Data is data is data
When we delve a little deeper into the requirements for working effectively with IoT data, things are not as simple as they may initially seem when reading various product and company presentations. E.g. there are many IoT platforms that aim to give their customers open access to their data. There is nothing wrong with that. In fact, it should be the least you can ask for.
“there are many IoT platforms that aim to give their customers open access to their data. There is nothing wrong with that. In fact, it should be the least you can ask for.”
A more important criterion is whether the data model that encapsulates the raw time-series and machine data and ensures that the raw data can be viewed in a business context is open and easily accessible. A further requirement for the data model is flexibility that allows raw data to be interpreted and put together in new contexts, as needed, and that raw data is not replicated every time a new data model is created. Another general requirement is that it must be straightforward to integrate to data models, live data and the IoT platform’s services. It requires an open architecture that can scale in terms of data volumes as well as the platform’s ability to handle data should scale effectively when more types of users and organizations are added. A final overarching requirement is that you should be able to guarantee that data and not least the data model (and thus probably the entire platform) is under the company’s control. In practice, this will mean that as an owner of an IoT product or as a data owner in an Industry 4.0 process, you should have the option to move the entire solution into your own IT infrastructure, which can be cloud or on-premise.
Let’s see how Beacon Tower addresses these requirements:
Beacon Tower uses an open architecture
Beacon Tower has an open architecture without the use of “black-box” functionality. At the same time, all services are exposed via APIs in a service layer, and thus the raw data from machines and sensors can be freely accessed, but even more important is that you can integrate directly with the services of the data models and thus reuse the data structures in other systems. It is e.g. incredibly useful in a machine learning context, that the data analysts can use the data models and do not have to spend time structuring the data. This means e.g. that a data analyst can query Beacon Tower about the following:
“Give me the Model X machines that have had an atypical vibration pattern within the last 24 hours for our customers in Scandinavia”
In this way, the data analyst can work effectively with e.g. to improve the anomaly detection algorithm in its development environment of choice. Likewise, the service technician can have a dashboard in e.g. CRM or its service application that shows which machines show signs of needing maintenance within the person’s area of responsibility.
Beacon Tower has a flexible data model
In order for the above data analysis scenario to be possible at all, it naturally requires that you can define your data exactly as you would like it. In Beacon Tower, you can define as many data models as you like and insert them into an infinite number of hierarchies. Therefore, it becomes easy to structure data on e.g. the following way which uses a company that produces, sells and services car chargers as an example:
Beacon Tower is repeatable and dedicated
Beacon Tower is designed to simplify and accelerate IoT development and operation by running the entire platform in the Azure environment of choice and, thus, is dedicated to the individual customer and their business partners. The entire platform is continuously updated and maintained. This removes much of the uncertainty and complexity of building reliable, scalable and secure IoT applications, while at any time you can have a development and test environment or even a dedicated solution for one specific customer. It also means that you have 100% control of the platform and can guarantee to customers and business partners that data is located in one specific place.
Read more about how Beacon Tower is a natural part of an integrated data platform in an enterprise IT environment.
For more information:
Partner, Mikael Rönde, firstname.lastname@example.org
Read more about the different Beacon Tower delivery options.