Blog

Basis of IoT: data, analytics, prediction and optimization

Extending the service interval of a paper machine, optimising energy distribution or improving household security systems, for example, will bring significant savings, which will make processes more efficient and enhance wellbeing. Data and analytics are the key to this.

The development of the Internet of Things (IoT) is bringing new types of business models. Conventional business models that focus on production and transactions are increasingly being replaced by models offering more extensive service packages. Closer cooperation is leading to longer customer relationships, and services are accounting for a larger share of total sales.

  • “At the initial stage, IoT is being developed in the industrial sector, but as the price of the technology comes down, smart solutions will very soon be a common sight in people’s homes as well,” says Managing Director Antti Syväniemi of Houston Analytics.

Collecting and analysing relevant information is increasingly at the heart of business operations in this digital age. This requires identifying and understanding not only the data sources for making fact-based decisions but also the information process for synchronizing the data to allow for integration with the organisation’s decision-making process, actions and monitoring. The more diverse and precise the data retrieved from the processes and the customer interface, the better the conditions will be for optimising the production, products and services.

Based on clear responsibilities

A diverse range of technology is needed for the two-way smart chains required by IoT and for data control and enhancement of the efficiency of decision-making. This technology includes measuring devices, routers, data storage, and analysis and optimisation tools. Syväniemi emphasizes that in addition to data processes and technologies, successful use of information requires a clear definition of responsibilities and authorisations within an organisation.

  • “Unclear responsibilities can easily lead to a confusing picture of the analytics practices and the systematic processing and distribution of data, which slows down the changes taking place in processes and culture – and it is these changes that are required to ensure good use is made of the data.”.

Once the responsibility issues, data collection and data synchronization have been solved, the tools to process the information can be addressed. In this era of data-driven management, the toolbox of each and every company should include explanatory and predictive analytics, optimisation and smart reporting tools. Analytics will reveal causalities, and this will improve the process of acquiring answers to questions. Various predictive models can show what will happen if operations continue unchanged and what the best and worst scenarios are. Such models can also be used to list alternative solutions and to produce predictions of the effects of different decisions on a ‘what if’ basis. Optimisation will indicate the choices to be made to achieve the desired goal.

Easy to begin

Companies that manage information flows and can develop and offer superior processes and services based on analytics and an operational vision will beat the competition, produce greater added value for customers and have the potential to prosper in the future. For a company to maintain a dynamic strategy and optimal operations, it is vital that it not only monitors the past but is also, crucially, able to forecast the future.

  • “The whole world need not be changed all at once. A good way to begin the journey towards data-driven management is to identify the most critical problem areas for processes and services – and to fix them first. Proceeding one step at a time with the aid of analytics will bring the best business dividends. The battle is half won if you have a good tool. For example the easy-to-use and easily integrated SPPS Modeler allows results to be applied quickly,” says Syväniemi.

Savings through service interval optimisation

There have been good experiences with analytics and predictive models, for example in the paper industry, where analysing the information provided by sensors has made it possible to extend the service interval of paper machine rollers by nearly 20 per cent. This means significant production savings and better reliability.

Better customer service through indoor positioning

The storage of indoor positioning data for retailers is becoming more efficient and affordable all the time, allowing quicker use of the data. Customers can receive additional information on the products they pass, on targeted offers, and on various services using a smart shopping trolley equipped with a tablet computer, or the customer’s own mobile device.

Availability or quality guarantee based on predictions

At bars and restaurants, the availability and freshness of beers with a short shelf life can be verified using sensors attached to the barrels. The sensors monitor the movements of the barrels inside the bar or restaurant or storage area. Scales underneath barrels from which beer is already being dispensed provide real-time information on the amount consumed. This data can be used to predict the availability of the beer and to indicate its freshness.