Using data science to raise the value of IoT data

Within the next three years, the number of IoT devices will pass 14.4 billion. With so many devices in operation, it might seem daunting to collect and use all this generated data, however, there are now IoT platforms available that can integrate multiple devices and present information easily. The real challenge for some IoT applications might be to make sense of such large quantities of data effectively; one possible solution to this problem would be harnessing a Business Intelligence platform and data science tools to maximize value by identifying correlations between different sets of data and running forecasts.

Why Data Science and why a Business Intelligence platform for IoT?

As the USDSI (United States Data Science Institute) has stated, “Data Science and the Internet of Things, i.e., IoT – the future of technology – are compared to each other – but the truth is they both complement each other.”

IoT applications usually generate thousands or millions of data points each and every day, and data science can offer an intelligent way of using such data. Therefore, the union between data science and IoT can benefit businesses in a variety of different ways.

Of course, implementing data science tools is not an easy task, so to facilitate this process, a Business Intelligence platform may be the best solution. For example, a BI platform can help you understand how devices are used over time to improve product designs, detect and diagnose problems automatically, develop new business models based on customer usage patterns, or even create predictive maintenance schedules. Those are just a few examples of how a BI platform can help an IoT application; find more use cases here.

Advantages that a Business intelligence platform brings to your IoT application

Besides the possible use cases, a BI platform will bring other advantages; we will go through some of them:

  • A Business Intelligence platform implemented in an IoT application can facilitate the extraction of value from your data and provide insights;
  • Machine learning and data analysis can help you make more informed decisions about your devices and data. For example, you can use these techniques to learn about your devices’ performance and what changes or improvements need to be made;
  • A business intelligence platform helps you use the info collected by your IoT implementation. With sensors and devices always creating data, it can be tough to make sense of it all. Therefore, the platform organizes the data and highlights trends and patterns you would otherwise miss to get more value out of your IoT implementation data;
  • The platform can also automate data collection and analysis, which will free up your staff to focus on other tasks. For example, if you need to generate a report regularly, you can schedule the report to be generated automatically. In addition, machine learning techniques can get more accurate results faster than trial and error methods.

Challenges you might face integrating a Business Intelligence platform

While a BI platform can free up your staff by automating data collection and analysis, initially integrating a BI platform can take time and effort. It’s not as simple as just installing the plugin and starting to use it. We have separated some of the other challenges that might come up along the way:

  • The platform might need some time to learn and understand all the different types of data being collected by the IoT devices;
  • Data interpretation is not always straightforward, and the platform might need to be configured to display the data in a way that is easy to understand;
  • The platform needs to offer enough flexibility to accommodate changes in the future as the company’s needs change and evolve;
  • It can be expensive, depending on your needed features. Usually, there are modules to choose from, including different features prepared for various necessities. The price may depend on how much data will be stored, the number of users, and how often you send or analyze data.

What to look for in a Business Intelligence platform

Ensuring that the platform you choose offers what your application needs is essential. Some things to look for include:

  • The ability to connect to a variety of data sources, including both internal and external data;
  • A highly performant data engine to handle large volumes of IoT data;
  • Ease of use, with a user interface that is easy to navigate and understand;
  • A wide range of analytical functions, including statistical analysis, forecasting, and data mining;
  • The ability to create custom reports and dashboards that show precisely the information you need;
  • To keep track of data versions and training.

The Domo plugin inside TagoCore

Based on the BI platform’s advantages, we are implementing new ways for developers to benefit from data science and business intelligence tools. TagoCore now allows the use of plugins to quickly and securely connect with robust external data science and business intelligence tools to make more sense of their data.

We have selected Domo for this first integration among some great providers that offer superior business intelligence and features. Domo is a cloud-based platform that provides various data science features, such as machine learning and predictive analytics. This new plugin for BI tools will allow TagoCore users to get more value from their data and make better decisions for their IoT implementations.

How it works

The Domo plugin for TagoCore can synchronize data between the two platforms. You can receive data from a Domo DataSet and insert the data into a TagoCore Device. You can also send data from a TagoCore device and insert it into a Domo DataSet.

The synchronization process will occur once every hour, and if there is no active internet connection during this period, the synchronization will wait for the next hour before synchronizing.

During the synchronization, TagoCore will check if there are new data points that need to be sent to Domo, and it will also fetch new data points from the Domo service that need to be inserted into a local Device.

Data science tool and IoT

The diagram above presents a scenario where the send and receive options in the Domo plugin are enabled. As you can see in the diagram, the flow starts with the plugin communicating with the TagoCore API to check if there are data points that need to be sent to Domo. The plugin will then make either one (1), or two (2) requests to the Domo API.

  1. If new data points need to be sent to Domo, TagoCore will send them.
  2. Then, regardless of new local data points, TagoCore always requests to fetch data from a Domo DataSet and insert new data points into a local device.

You may disable the plugin’s ability to send or receive data by toggling the switch in the plugin’s configuration.

Keep in mind, however, that new data points fetched from the Domo service will always follow the usual insertion flow, meaning that they will be parsed by a payload parser and an encoder module if one is available before being inserted into the bucket.


We can use our TagoCore Freezer Simulator plugin as an example of an application. Once this plugin is downloaded and activated, it will constantly send data simulating an actual freezer to TagoCore. You can customize the temperature scale, frequency of data, and even to which device to send data.

This plugin is ideal as an example because it represents a genuine use case and can even represent other scenarios where you must constantly send data and synchronize it with Domo.

To install the TagoCore Freezer Simulator plugin, head over to our Plugin Store inside of TagoCore.

Installing the Domo plugin

To use the Domo plugin in TagoCore, you just need to install it. To install the plugin, head over to the Plugin Store by clicking on the Store icon in TagoCore. Once the Plugin Store opens, locate the Domo Integration plugin and install it.

Adding the credentials

Once the plugin is installed, you are one step closer to integrating your Domo account with TagoCore. The next step is to add your Domo credentials so that TagoCore can make requests on your behalf. To add your Domo credentials, you need to create a new client in Domo and add the Client ID and Client Secret of this new client in the plugin settings of TagoCore.

Sending Data to a Domo DataSet

To insert TagoCore data into a Domo DataSet, you need to enable the Send data to a Domo DataSet option. Once you enable the option, you must inform which devices to acquire data from (single or multiple), as the image below shows.

Business intelligence platform implementation with TagoCore

Selecting data from a single device allows you to choose a Device ID for the origin of the data. Selecting data from multiple devices requires you to inform the tag key and tag value of the group of devices.

By sending data, TagoCore automatically creates a new Domo DataSet with a specific set of columns: Variablevalueunit, and time.

Receiving Data from a Domo DataSet

To insert Domo data into a TagoCore device, you need to enable the Receive data from a Domo DataSet option. Once you enable the option, you must inform the desired DataSet ID in Domo and the Device ID in TagoCore.

In conclusion

Data science enhances IoT by providing an insightful way to study the extensive data collected. In certain scenarios, data science is even crucial for the success of an application. It’s safe to say that the benefits provided by data science surpass the disadvantages; even though some challenges may appear during the implementation, they will most likely be worth it in the end.

The plugin is another way to use data science with TagoCore since someone could do it by themselves using our analysis mode. Furthermore, this plugin is just one option from the many plugins that TagoCore offers; it’s possible to personalize your application as you wish from the database you prefer and even create your own plugins.