This is the second part of the series Fields of Application of Big Data. In the first part, we have discussed the basics. We are starting this part discussing the marketability of Big Data.
It can give the impression that the results of big data are equally common in every field of application and weighted equally. However, all fields of application are diverse, so that in some areas specific aspects have to be considered more. The principle of protection of personal data must always be preserved, ie transparency must be respected, everyone must know that data are collected about him/her. If there is no transparency, an explicit consent of the persons is required. Furthermore, marketability is limited by any laws and risks in some fields. A factor for a possibly smaller market suitability are the respective nationally different regulations. In the case of data collection that does not comply with national regulations, uncertainty does not protect against the consequences of an infringement. There is a big potential for big data in marketing, but there is also some social risk due to ethical issues. Personalized advertising will be sent out, suggesting a familiarity to potential buyers. The potential buyer can take this negative because his private preferences are already known. Intrusion into privacy can create distrust and negatively impact customer loyalty.
In addition, it must be ensured that suitable data is always analyzed on the market. Records must be complete to ensure a representative analysis. If there are isolated or even large amounts of incomplete data, this can lead to problems in the analysis and also in the validity of the evaluation. In its study from 2014, Bitkom states that complex relationships can be shown professionally and without errors, but at the expense of easy-to-understand presentation. The consequences here can be error interpretation by third parties, It can also be too time-consuming for SMEs to carry out huge analyzes. Possibly. Large technical systems must be purchased, which would result in immense initial costs. Here it can happen that the profits can not cover by the surplus value, the costs of the new acquisitions. Thus, an acquisition of such systems is not economically to argue.
These are just a few short examples that give an overview of the fact that big data can not be used in the same way in every field of application and there are also some restrictions in certain market segments. Only a very small proportion of entrepreneurs and employees believe that there are no implementation barriers for big data.
Business Life Application
Research and product development
Over 50% of entrepreneurs use big data to develop new products and services. Businesses continue to see big benefits in big data in terms of product development. Relevant focal points of big data mechanisms in research and product development are new product ideas that emerge through analysis of social media or the problematic introduction of new medicines in the health industry. The research not only aims to generate new product ideas, but also tries to improve existing products on the market. By including social media sources, the products can be constantly evolving based on the consumer’s criticism. This method is also called crowdsourcing.
A well-placed product requires previously created competitive advantages and an efficient time-to-market strategy. Such added value can arise through the use of product prototypes. This practice already existed before big data and is better known as rapid prototyping. Evaluations of big data help to produce more efficient results.
Each new prototype is build incrementally on the insights gained from the previous prototype. This is how a product approaches perfection. The prototypes developed are characterized by high flexibility and ensure automated production by the evaluation of data. Such prototypes can be used in test scenarios to reflect personalized products that are currently in great demand in the automotive industry. This appeals to customers who value individual products. On the one hand, there is a standardized construction kit for cost reduction, on the other hand, the test prototypes are designed to produce recognition values.
Due to the continuous development and the integration of prototypes, the automotive industry has to record an increasing product quality in recent years. Even if the findings from data analysis can lead to exclusivity and therefore competitive advantages, they are available on an industrial scale within a very short time. Therefore, constant further development is necessary to be able to present unique new products. Big Data helps to gather action-related statements regarding research and development issues. The result of their use in research is therefore individualized offers and optimized product development.
Electromobility, Underwater Robotics or Logistic Robotics Systems are just some of the fields where Big Data can be used for development. Especially in the field of electromobility, the improved collection of vehicle data is being researched. As human-type robotic systems are also being developed, the use of sensors plays an important role. Data and motion sequences can be captured by sensors and transmitted from human to robot systems. Such sensors generate a variety of data. To summarize, the existence of the data and its evaluations creates transparency that can have a positive impact on new products developed, inter alia, by research.
Financial and Risk Controlling
Controlling is an important instrument of the higher management level. In order to capture all data properly, it is necessary today to support information systems. Thus, the data is collected and evaluated using Big Data analysis systems. The required functions are increasingly being integrated into CRM systems as well as ERP systems. Currently, most companies rely on KPIs for planning and control, which they need to examine in terms of corporate strategy.
Reporting and analysis are already taking place using these KPIs, and Big Data is adding additional metrics to these methods. In contrast to the former reactive behavior of the controllers, big data adds further possibilities, which include a proactive behavior. Controllers have the opportunity to perform more reliable simulations from the data obtained. Scenario analyzes can also be presented in a more granular and secure manner, thus better assessing business cases and market potential. Big Data evaluations benefit Controlling with three significant added values in the company:
monitoring “What is happening now?”
Predictive “What will happen?”
Presriptive “What should happen?”
However, the integration of big data into controlling processes can not be implemented without the creation of basic prerequisites. Good corporate governance reduces the risk of a wrong decision through novel Big Data insights. The models and analyzes used in controlling must be clearly known, flexible and dynamically adaptable. The change to big data evaluations has to be carried out professionally by the staff and therefore requires appropriate training. By using relevant information, controllers can work more efficiently to guarantee up-to-date technologies, the highest possible up-to-dateness, and the widest possible relevance of reported information without generating information overload.
The evaluations by machines are gaining importance. However, this increase does not mean that the significance of the factor human in the area of controlling decreases. Controlling will become more important in times of big data and postpone a shift in the main tasks. The controllers are less concerned with collecting data than with analyzing the collected data. In Financial Controlling, a higher level of automation can be reported with significantly less effort. The degree of information provided by these data is then filtered out by the “human controllers” and the resulting added value for their company.
Today’s production processes are characterized by a high division of labor. This division of labor is characterized by diverse machines and cooperation between companies. The emerging data volumes are a major challenge for the control of production. Data is often redundant across companies and is evaluated differently. Data protection also plays a role here, as companies are reluctant to give their own data out of their own homes. Therefore, companies are now specializing in the development of appropriate tools to increase the availability of production machines. It focuses on the use of adaptive data mining methods that individually calculate the optimal production process for each manufacturing process.
In steel production, one third of the steel produced annually is pure scrap. The resulting steel scrap also passes through the production lines and blocks the machines during this time, which is very inefficient. Big data analytics should evaluate the production data, modify it and thereby produce more efficiently, which saves money. There are projects funded by the Government, which deals with Big Data evaluations in production processes. Not only theoretical analyzes but also practical processes in real production are simulated. The surface of the materials is checked for unevenness and sensors are used to monitor vibrations and temperatures. After evaluation, faulty materials are detected directly and do not have to go through the entire production process. The quality assurance for the parts produced can be done in advance by machines, so that the human worker can be alerted to any errors and less time needed to find these errors. Likewise, entire human quality checks by such mechanical mechanisms can be omitted. Here, too, big data not only helps with the analysis of the actual state. Optimal business processes should be forecasted and automated.
So, in addition to the reactive behavior, it is possible to make forecasts for the future. The past material errors can be analyzed, and then to be able to say which errors are most likely to recur in the future. It can be predicted a defect of a machine part in advance. Thus, the premature replacement of a future defective machine part is guaranteed, before it really fails and so the entire production faltering. This premature treatment of future errors is called Predictive Maintenance.
The big data analyzes that have been applied to individual production machines up to now are increasingly being linked to other production machines. This increases the complexity of the data. The components of the products come from many manufacturers, so that the data must be evaluated comprehensively in order to ensure a steady production improvement. In the value added of a product, the importance of the data increases enormously, so that in addition to capital, labor and raw material the resulting data can be seen as the fourth factor of production.
Conclusion of this part
In this part, we have discussed about some topics which can be put under the header “Business Life Application”. They were (i) Research and product development (ii) Financial and Risk Controlling (iii) Production. In the next part, we will complete rest of the topics which can be put under the header “Business Life Application” which includes (iv) Marketing and Sales and (v) Distribution and logistics.