See the original Interview in German about Industry 4.0
Produktion asked Dr Brunner about his thoughts as to why data analytics is the base of "Industry 4.0".
Dr Brunner: Industry 4.0 is based on many new concepts that cover connectivity and intelligence generated from data. Examples are machine-, sensor- or log-data. That’s why one of the challenges companies face is the huge increase in data growth: Soon we’ll face multiple zettabytes of data to gain insights from. Additionally, analytics get more efficient and are performed in real-time. Examples are real-time predictions and improving production processes. Next to increasing speed of data growth, data variety increases as well. Those different types of data must be combined, analysed and combined again with structured data like ERP or CRM.
Dr Brunner: There are different benefits from big data. One way to think of it is in terms of cost reduction. This goal can be reached by preventive maintenance for example. Another way to think of it is the creation of completely new business cases. One example is Adidas which now produces individualized shoes for their customers. Those use-cases can be created and/or supported by machine-learning-algorithms like recommendations, classifications and clustering. By now big data has overcome its “hype”-phase. Companies started to adjust their big amounts of data to finally run machine-learning-algorithms on it. From now on the company with the fastest analytics, best data and best future planning will have the biggest competitive advantage.
Dr Brunner: For quite some time developers of big-data-software postulated that it is enough to install the software and start collecting and analysing data. To me, that is not useful, because money is often paid for licensees instead of valuable business insights. Sometimes different systems are bought, which creates a zoo of different software. Everyone choosing this way might face problems when it comes to finding experts for special solutions. Instead of that, I recommend to choose one specific problem or business case and start with one goal. In most cases one system is sufficient. The most important requirement is to define a process model with adequate architecture. That is a great basis to build up from. Additionally, data quality is important. Normalising and adjusting is a time-intense process that is sometimes underestimated.
Dr Brunner: From our training, we see that industry 4.0 teams are very multidisciplinary. Next to specialized departments, the IT team, and analysts, there are members of security and data protection. Especially for production plants, the IT process is complex and many departments are connected, each with individual challenges. Therefore, a training is a very good way to bring everyone together and onto the same level. Thereby a shared strategy can be developed. In case a specialized department wants to answer an individual question the training can be adapted to it. It is important, that interest and openness towards new digital topics are given.
Dr Brunner: We work with practical examples from experience and participants can work on use-cases themselves. Additionally, we look at system demonstrations. Usually participants are surprised by the huge variety of possible applications occurring in big data surroundings. Often a clear vision of the individual architecture is the result of training.
We are now working with online courses, coding challenges and coding coaching. We can fulfill the needs of a very large community of data scientitsts to be trained on how to analyse and handle big data in Intrustry 4.0.