16. MÄRZ 2018
Industrie 4.0 is about process development
A transition to a data driven company does not require vast amounts of money,
but rather an approriate process strategy. This was a key message at the ‚Industrie 4.0‘ forum in Munich organised by the international consulting firm 29FORWARD AG.
The importance of Business Intelligence in common production centred companies increases as the volume of data required by its business processes growths. As a result, German based companies call for more data analytics capabilities to develop and optimise their processes. Many wonder what specific measures would be necessary to tackle the challenges of a comprehensive digitalisation. To answer this, 29FORWARD, a consultancy firm specialising in Business Intelligence, organised a forum in Munich to discuss Industrie 4.0 with industry specialists from relevant companies. Real-world projects were presented which demonstrated the current level of sophistication in the industry.
For Ralf Neufang, CEO of 29FORWARD AG, and for many industry leaders, Industrie 4.0 represents the fourth revolution in the business, which centres around the re-design of processes within a company as well as industrywide. However, not all companies are taking steps towards Industrie 4.0. „According to a recent survey, only 50% of companies affected have taken measures towards Industrie 4.0.“ says Neufang, „This, however begs the question why the other half have not been active towards Industrie 4.0. There must be plausible reasons for this.“
Process optimisation through Business Intelligence
„Many projects are simply too big, dictated by management and not well prepared“, which are the common reasons why many projects are destined to fail according to Neufang.
Andreas Raaz, partner at 29FORWARD AG, explains in his presentation why a transformation to digitalising processes has to be tackled in smaller increments. Showcasing a project for a european airport with over 40 million passengers a year, Andreas explained how the Proof of Concept method was the ideal choice: „This approach allowed decision makers to gain insight into the company‘s own data and to assess realistically the efforts that were required and the added value that could be generated.“
Raaz also noted that data management represented an important part of business analytics in the competition for market share – with data quality and value being of significantly higher importance than the quantity of information. BI consultants Peter Fuchs and Marinko Jurcevic agreed and added that many technologial terms around data analysis were being used incorrectly. They advised companies to be cautious in the use of new technologies and warned not to follow new hypes blindly.
They refered to neural networks that are increasingly used by firms as an alternative to classical methods, such as decision trees and logistic supply chains. According to Fuchs however classical methods are still very relevant today and have the advantage that they are easier to interpret and more straightforward to implement, while they also tend to remain more robust in practice.
Neural networks are only part of the solution
Based on an example Fuchs demonstrated how the use of neural networks and algorithms used in the field of AI can be highly useful for Business Intelligence. In only a few weeks he created a robust photo recognition tool using simple coding methods and the free software Tensor Flow. For the purpose of this exercise the tool differentiated between cats and dogs, however the code could be amended as required so the tool could be applied for quality control within the production process.
Dealing with with large volumes of data is also a challenge for other parts of the business as Friedrich Kitzsteiner from Carl Zeiss AG pointed out. Last year the company entered into a co-operation with Cisco in the field of global machine connectivity, with the aim to develop a secure data transferring system to enable Zeiss devices to be linked to higher order systems within Industrie 4.0.
During the discussion however he talked about his company’s digitalisation project for Office Services which involves electronic digitalisation and distribution of all corresponadence including magazines and letters.
Define the process first, then digitalise
Dr Frank Sinka of OSRAM GmbH also demonstrated in his presentation how networks trickle through to the daily business and generate agile dynamic changes within the organisation. Data dynamics is an important part of the Industrie 4.0 strategy as well as for the ‚organisational structures of the businesses of tomorrow‘. As Steven Doempke from Sigma Chemnitz GmbH pointed out the knowledge around company data is also crucial. His presentation focused on contactless identification and localisation of objects. The data collected would be of value for the fields of logistics as well as inventory security.
Participants agreed that prior to implementing new technologies the corresponding processes had to be well defined. How this could be achieved for the projects presented formed part of the closing discussion. 29FORWARD lead this discussion to get an understanding of the sophistication of current Business Intelligence projects as well as to provide support in finding solutions for implementation. Problems during implementation can arise when processes are not sufficiently analysed prior to implementing Business Intelligence. Another common reason for failure are promises of big projects by software developers that are too big from the outset. As Neufang knows, relying on software to do everything as long as it is being fed lots of data usually ends in disaster. He advises to plan big but start small. In his view prototyping is one of the most invaluable approaches.
Images: © Andreas Burkert