05.04.2018Blogg

Visualisation Provides Better Understanding of IoT Data

The Internet of Things (IoT) generates a huge amount of data (Big Data), which is useless in itself, unless it is converted into a format that is easy to understand, analyse and present. Merely connecting IoT devices and collecting data is no longer an achievement. Instead, what matters is how you utilise the IoT data in your business. I claim that visualisation may be a key part of the IoT data utilisation process.

People often believe that visualisation is just a collection of graphs and infographics. However, visualisation requires a basis formed by the analysis and understanding of data, and most of all by the definition of needs and applications. In this blog post, I explain how you can convert IoT data into a useful format to gain information that will benefit your business.

If you simplify things to some extent, you can say that any IoT data is data of the same type, regardless of the values measured by the sensors. Data is generated at specific intervals. The interval may be anything from milliseconds to hours or even days, depending on the required measuring interval or susceptibility to changes. In addition, a specific variable or several variables are always measured.

Monitoring provides good visibility of data

The data from a sensor can be very easily and intuitively visualised into a time series using a traditional line graph, for example. There is a place and need for such visualisation of the data generated by an individual sensor.

The most common reason why the data generated by a single sensor needs to be visualised is probably to monitor the status, operation or location of a device, or at the beginning of a project simply to make the data visible.

IoT data imposes some requirements on the technology of the visualisation tool used. In the case of monitoring-type visualisation, real time data is an absolute requirement. Another clear requirement involves the volume of data: the technology must enable the visualisation, aggregation and filtering of millions of data points.

From active to passive monitoring

In an optimal situation, visualisations that are created for monitoring needs become less significant as the project proceeds and matures. Limit values and alarm limits can be specified based on an analysis of the IoT data or purely based on the needs, which means that the focus switches from active monitoring to passive monitoring where targeted input can be sent based on the data whenever specific measures are needed.

The input, alarms and specific rules improve operational efficiency and reduce the time used for manual monitoring.

The difference between monitoring and alarms can be illustrated with a simple example regarding the refrigeration devices in supermarkets. The temperature in the refrigerators should remain between specific limit values. Instead of manually checking the temperatures by physically going to each refrigerator daily or more often, monitoring can be used to check the temperatures from one screen at specified intervals.

The optimal case is that a device will issue an alarm if the value remains below or exceeds the set limit value.

The alarm is shown in the terminal device of a person who will be able to take the necessary action in order to prevent the loss of products, to verify the condition of the device and to have it repaired, if necessary. Sometimes the person only needs to go to the refrigerator to close a door that a customer has left open.

Visual analysis enables the development of operations

The great benefits offered by IoT data cannot be achieved directly by visualising or monitoring the IoT data, however. Instead, the benefits can be achieved by analysing data from devices and sensors and using the IoT data to derive indicators that allow for control of the operations to improve the business, to reduce energy consumption or to make the operations more environmentally friendly, for example.

The benefits can be achieved by analysing data from devices and sensors and using the IoT data to derive business indicators.

Sticking with the example of refrigerators in a supermarket, you can easily calculate variance for the fluctuation of temperature inside a refrigerator. The lower the variance, the less energy the device should consume. You can set a target level for the variance and study the differences between the variances of refrigerators, for example.

IoT data can enrich your organisation’s other data

In most cases, the ultimate benefits can be achieved by combining IoT data and the indicators derived from it with the organisation’s other data. Let’s remain at the supermarket: you install sensors on each aisle and at the entrance that provide information on how many people visit each aisle. The number of visitors is already an indicator which describes how many visitors there are at a specific time of day, why people come to the store, what is the most interesting aisle, how many people in total visit the cosmetics aisle, etc. But what if you combine the data on the number of visitors with sales data? You could easily calculate how many of the customers who visited a specific aisle actually bought something and how much money they spent. Is there an aisle that people frequent without buying anything?

By combining data, you can obtain whole new information about the business, which allows you to improve your operations.

Merely connecting IoT devices and collecting data is no longer an achievement. Instead, what matters is how you utilise the IoT data in your business. You can use IoT data, particularly if you combine it with already existing data, to discover new business patterns and trends. You can use the data to develop your operations, optimise costs or even create new services and cash flow.

Jenni Linna works at Solita in a team that specialises in IoT data. She focuses on the visualisation of data. She is passionate about data analysis and understanding data, the definition of customer needs and the creation of visually appealing and relevant dashboards.