A fresh mix of data creates new insights

Although big data as a concept signifies a vast and abstract mass of potential solutions across industries, Mark van Rijmenam, from the big data platform Datafloq, advocates increasing the use of the term mixed data instead. He claims that companies can get the most out of data by combining different data sets, not only from their own processes, but also from external and public sources.

Choosing an optimal data-driven approach is a promising but tricky challenge for companies. Many question what data source is best for a particular process and how should the data be analyzed. In addition to employing and mixing a company’s own data resources, Mark van Rijmenam sees possibilities in going one step further and looking for useful data beyond company borders.

“Combining data sources, what I call mixed data, is becoming more common as many companies are beginning to combine their internal and external data sources. If a company has detailed information on its customers and external macro-factors that influence demand, be it geopolitical news or market crises, it is able to optimize its inventory further”, van Rijmenam explains.

Van Rijmenam continues by highlighting one consumer business company, which is already proceeding full-steam with mixed data, and could provide a model for companies in the manufacturing business as well.

“One of the best examples of a company utilizing mixed data is Walmart. They have been able to optimize their up-to-the-minute inventory per location based on all kinds of internal and external data sources. This method is completely scalable to all industries: if you have data on a product that you have developed and it’s compiled of different raw materials, you already have a few useful data sources at your disposal. Any company can use similar data sets to further optimize their inventory on the basis of, for example, the price level of raw materials, demand or weather conditions.”

This type of data mixing, which draws and combines data from different sources is already in use in a specific area of energy production, which is most susceptible to the forces of nature.

“A Swedish wind energy company uses big data to find the right spots for the mills by analyzing wind and current patterns and temperature over the sea and connecting this to data regarding energy prices. They have also used predictive maintenance to get rid of the need to fly out with a helicopter every once in a while to see what’s going on. The sensors tell the company when and what kind of maintenance is needed.”

In order to reach their full potential in all industries, the industrial internet and big data analysis demand the right technology. Van Rijmenam considers the level of technology to be advancing at a sufficient rate to make this progress happen in various industries.

“Many different start-ups and long-running companies are currently developing smart algorithms by using machine learning and artificial intelligence to achieve deep insight by combining different data sets. During the next few years we will see an exponential growth in the development of smart algorithms, as more companies will dive into this field. We’re almost reaching a tipping point, as more and more industries are realizing the importance of big data and the industrial internet, and start raising the demand for the right technology.”

Key requirements and benefits of a data-driven culture

According to Mark van Rijmenam, a company that wants to start applying big data in its processes needs to implement a new type of thinking across the whole organization.

“In order to work with big data, an organization, including the board, has to have a shared understanding of what big data is and what it can be used for. The result of this should be a data-driven culture, where each department relies on the data that is being analyzed and uses that insight as the basis for decision-making. It should also be ensured that the data and the algorithms created are reliable and correct.”

Van Rijmenam notes that the benefits of exploring your business through big data are available to all industries.

“Certain benefits apply across all industries: getting to know your customers better, recognizing your risks and optimizing your inventory or operation. For example, in process industries involving material handling, big data helps companies to monitor their products and equipment better. Sensors allow you to know exactly where everything is. By combining this insight with data on how employees work or how the products are used, companies can optimize safety and improve efficiency in their inventory.”

“In the steel industry, a company can find ways to improve its margin by gaining a profound understanding of what is happening in the manufacturing process of steel. This can be achieved, for example, by running simulations and combining different data sets to find out how the plant operates and where its different components come from. The company can discover hidden bottlenecks and anomalies in the processes by breaking them apart with data analysis. Taking advantage of this new insight, and therefore improving operational processes can result in millions worth of savings each year.”

Instead of starting off with applying a data approach to the whole business process all at once, van Rijmenam points out that a company should begin with a process that is easier to grasp and can be tested by using a proof of concept.

“Formulating a proof of concept is a very important aspect in getting started with big data, because developing a data-driven strategy and implementing it is not an easy job. Typical big data projects last approximately eighteen months, so you should start small by learning and experimenting on what it can do for your organization. When you have done that, then you can expand.”

Data marketplaces on the rise

As reliable and accurate data and the ability to take advantage of it is becoming a valuable resource for many companies, a new phenomenon called big data marketplaces is emerging as a result. Mark van Rijmenam shares his views on this development.

“In the coming years we will see more marketplaces for big data, where companies can either publish their own data or find data from other companies. This will allow companies to access better data sources and, through mixed data, use it to acquire new insight. In addition to the already existing marketplaces, an increasing number of governments are opening up their data sets. I have been in touch with a company from Singapore called DatastreamX, which is developing a marketplace for real-time data. More and more data is being shared among companies in order for everybody to benefit from it and I think that is a fascinating development.”

Currently big data is generated through all sorts of processes from companies’ own processes to freely available sources such as scientific satellites. Could all of this data be valued somehow and be considered as a trade currency? Van Rijmenam discourages such thinking and stresses the relevance of data only as a means to an end.

“Naturally data should meet certain standards and regulations, but what is valuable to you might not be valuable to me. There is no point in putting value to data, it should rather be directed towards enhancing the analytical capabilities that you have. Data in itself is useless, it’s the insight that you can derive from it that counts.”

Mark van Rijmenam is Founder of Datafloq, a Big Data Strategist and author of the book Think Bigger – Developing a Successful Big Data Strategy for your Business

Image credit: Maksim Kabakou/Shutterstock.com