Digital Marketplaces Unleashed
Page 75
Although the changes in the process steps seems to be rather minor, the impact of these changes through automation is huge: cost savings and improvement of customer convenience by reduced service time: automatic spare parts ordering, van loading and visit scheduling
increased first implementation rate
same day appointments, if there is no spare part needed
decision support: decision makers can concentrate on complex repair cases
Looking at the Digital Navigator, the transformation has been conducted in the area of Digital Information Management (s. Fig. 48.3), especially in the context of Information Sourcing, Value Assessment, Analytics and Processing.
Fig. 48.3Digital Information Management
In Information Sourcing relevant data sources, in this case the already existing service calls, are identified and checked by significance and relevance through the digital capability Value Assessment.
Through Analytics a mapping is done to check, if the root cause of the current call can be solved in the same manner as an existing one. After collecting and analyzing the data, the call can be processed and the technician does have a lot more information about the scheduled customer visit.
48.3 Transformation Completion – New Horizons of Data
The situation is clear: the service use case described in the former section is completed. Really? How to check, if the ambition level has been as high as appropriate? Let’s take the Digital Navigator from section one to double check the use case, finding out higher ambitions.
If we identify the functional capabilities addressed in the use case and map them to the digital navigator, the result is surprisingly, but clear. We identify mainly capabilities out of the digital information management domain: Data Life Cycle Management: Basic to decide, what data and structures should be used and to care on changes on them and their environment
Information Processing: applying and processing the data for the purpose of the use case
Information Quality Management to ensure, that the applied data are relevant and in sufficient quality
Privacy Management, mission critical to ensure the distinguished use and integrity of the data
Analytics, to create the best combination of service parts using the historical data in an intelligent way
Presentation Exposure to offer the helpful results to all intended users, may be cross different devices
Great. Good coverage – of the information management domain only, to be precise. The opportunities of the information management domain are very well explored and applied. But, what about the other domains of digitization capabilities?
Remember – the story of digitization is about using data to create value. Of course, the first steps are about all the data, which are already available. Often, this is already a lot. Many companies are wondering, what to do with all the data, they collect and store into their databases. This companies should apply data strategies and business model strategies on their data as shown in our use case so far.
Anyway, in digitization this vision is not big enough. You have to care as well on a kind of data, which is not yet available. The Digital Navigator helps to find out. As an example we apply the functional domain Cyber Physical Systems (CPS) (s. Fig. 48.4). What is a cyber‐physical system within the context of our service use case? Or: What asset could become a cyber‐physical system, if it becomes connected? The basic idea in the phase of ideation is: Every asset coming along with the use case could become a cyber‐physical one. This includes: the washing machine itself
the storage
the spare part
the service technician
the service car
Fig. 48.4Cyber Physical Systems
Creating a cyber‐physical system out of them means, give them and their functions a kind of IT frame, which means represent their functions and data in IT structures – and connect them by an appropriate network. This includes fix line options as well as mobile networks including Bluetooth, WLAN, GSM, LTE or in the near future 5 G. At least they need an API as interface, so they can get a network connection and an appropriate IT integration.
The functional capabilities of the “Cyber Physical Systems” domain of the Digital Navigator give an overview, what is necessary to drive cyber physical systems. Some examples: Operational Integration of CPS: core to make it run
Real time Asset Management and …
… Real time Control and business monitoring: Both may be most important, because real time transparency is one of the core value driver
Event Driven Architecture: to make most use out of real time information, the IT Landscape need areas able to handle events in a valuable manner
Information Exchange, a difficult, but mission critical capabilities regarding different “languages” and standards in the communication between different assets
Most of these capabilities to use CPS today are pre‐integrated to Internet‐of‐Things (IoT) – Platforms, often flexible and secure offered by a cloud provider. The Digital Navigator helps to check, whether all needed capabilities are available in the IoT‐Platform of choice.
Applied to the use case, a lot of further opportunities are created immediately: 1.The washing machine gets a kind of self‐awareness, knowing about its condition and sharing this condition online with the service center. Transparency occurs on the concrete problems and therefore on the needed spare parts. In a next level, applying intelligence to the condition monitoring, the machine is able on self‐diagnoses in advance, prediction maintenance becomes possible, a new business model occurs: guarantee of the availability of the machine by the producer, a value which might be paid by users, especially professional ones (see Figs. 48.1, 48.2, 48.3 and 48.4)
2.The storage offers transparency on parts, which are available or which are not, where they are and when the next charge of missing parts will arrive. Linked to the messages of machine problems and linked to service orders it could become a self‐organized storage with close to 100% guarantee, that needed parts are available and the storage size is as small as possible.
3.Even the spare part itself could contribute to transparency and service, especially the bigger ones. Are they in an appropriate condition? Are they available at the expected place and time in the required type or version? What is the amount of parts on the desired service car?
4.The service car itself is the most interesting asset unused within this use case. It moves around and transport the required parts. Therefore, it could know in advance, if all required and expected parts are (still) on board to avoid customer visits because of missing spare parts. Moreover, it’s positon is known every time. Combining this with customer addresses and real‐time (!) traffic information, estimated times of arrival could be offered to customers very precisely, even online by a customer app or a web portal. A new service level of very small time windows could be offered to the customers. And more: Knowing every time the location of each service car and it’s available parts, it can be redirected by customer calls every time. A new kind of service can be offered: a same day service, in cities may be a same hour premium service.
An example deep dive into the service process shows, that a connected and self‐aware washing machine already changes a lot (s. Fig. 48.5). First, the machine user is neither surprised nor disturbed by an unexpected defect, if the predictive analytics detects the problem in advance. Second, the machine will not call a call center agent, it informs directly the service dispatcher. You can even imagine, the machine informs directly the service car closest to the location, because it knows, that the needed spare parts are on board. Customer’s will learn to be happy about the service technician knocking on the door, friendly announcing to care on the risky machine. In a near future, customers will expect it like this, never have to call
again.
Fig. 48.5CPS transformed process
In consequence, Digital Transformation obviously was not completed only using the already available data. Applying the Digital Navigator, it was pointed out clearly, that there were a lot of further opportunities out there, which means at the end: business opportunities. In our example this included a new business model on guarantee of availability,
a second new business model on express service,
an optimization of storage
a decrease of double customer visits increasing the one touch repair rate
Not too bad for applying only one more domain of the Digital Navigator. Valuating the results from the perspective of the current business model, taking into account, that the business model could be changed, a strategic roadmap could be constructed, where the opportunities are taken into reality step by step. Real data driven value is created using existing data, and, may be even more important, further data, which can be created by new IT capabilities and new connections on so far unconnected assets. Valuable transparency occurs immediately, driven to much higher value by using real‐time data and intelligence in analytics.
Embedding the roadmap into an overall digitization strategy of your company, all of the other domains of Digital Navigator help to identify, which current business functions have to change, or, what new ones should be established. Risk management for example gets an additional perspective. It becomes obvious, that a Digital Transformation Strategy touches wide areas of the company.
This is real digitization – initiated by the Digital Navigator
References
1.
Cross‐Business‐Architecture, “Digital Navigator,” [Online]. Available: http://www.cba-lab.de. [Accessed 2016].
2.
Detecon International GmbH, “Studie Digital Navigator,” Detecon, [Online]. Available: http://www.detecon.com/de/Publikationen/digital-navigator. [Accessed 01 09 2016].
© Springer-Verlag GmbH Germany 2018
Claudia Linnhoff-Popien, Ralf Schneider and Michael Zaddach (eds.)Digital Marketplaces Unleashedhttps://doi.org/10.1007/978-3-662-49275-8_49
49. The Data Science Lab at LMU Munich: Leveraging Knowledge Transfer, Implementing Collaborative Projects, and Promoting Future Data Science Talents
Thomas Seidl1 , Peer Kröger1 , Tobias Emrich1 , Matthias Schubert1 , Gregor Jossé1 and Florian Richter1
(1)Ludwig-Maximilians-Universität München, Munich, Germany
Thomas Seidl (Corresponding author)
Email: seidl@dbs.ifi.lmu.de
Peer Kröger
Email: kroeger@dbs.ifi.lmu.de
Tobias Emrich
Email: emrich@dbs.ifi.lmu.de
Matthias Schubert
Email: schubert@dbs.ifi.lmu.de
Gregor Jossé
Email: josse@dbs.ifi.lmu.de
Florian Richter
Email: richter@dbs.ifi.lmu.de
49.1 Introduction
The digitalization of the world is an ongoing process, influencing all parts of everyday life. The search for valuable information in growing amounts of data is undoubtedly one of the major challenges in the upcoming decades to gain the full potentials out of this development. The key to transferring data to knowledge is Data science, an interdisciplinary field that deals with algorithms and systems to automatically detect patterns that allows to derive knowledge from data in various forms. It brings together statistics, machine learning, data mining, and data management, extending the well‐established field Knowledge Discovery in Databases (KDD) by a more holistic approach.
Even though there is a high demand in know‐how for companies, thorough expertise and talents are still rare to find. For example, the well‐known report by McKinsey & Company projects a global excess demand of 1.5 million new data scientists. However, at least in Germany, there are only rare internationally visible research groups at universities and, thus, only few Master Programs with a strong and solid focus on Data Science are currently offered.
The Data Science Lab of the Ludwig‐Maximilians University Munich (DSL@LMU) is an institution for research, development and education around the field of Data Science. It is attached to the Database Systems group of LMU that has a long standing tradition in research in the fields of Databases and Data Mining and is constantly among the top‐10 leading research institutions world‐wide in these fields. In addition, the DSL@LMU has tight connections to the Institute of Statistics at LMU and other Informatics groups such as the group for Mobile Computing. As a consequence, the DSL@LMU provides the expertise of leading scientists in the key research fields of Data Science.
The key idea of the DSL@LMU is to link this scientific know‐how to Industry and Data Science students (see Fig. 49.1). The focus of the Lab is to bridge the gap from academia to industry and allow students and researchers to work on real‐world Data Science challenges while giving their industry partners a substantial leap forward. Besides, the Data Science Lab offers a wide variety of activities for small, medium to large‐scale companies that include professional training, trend flashes, summer schools, lecture series from international luminaries, workshops, guest lectures from industry, and many more. Thus, the top talents are linked to their potential future employers. A dedicated lab room gives the students the opportunity to work on the newest hard‐ and software products and at the same time offers an inspiring environment to discuss and elaborate new Data Science challenges. This room is also used for hands‐on professional training in Data Science related courses.
Fig. 49.1The DSL@LMU aims at bringing together the three groups including industry, students and academia in the field of Data Science
The Lab also fosters strong ties to the new elite program “Master of Data Science” (see below). Even though LMU Master students that have a focus on Data Analysis lectures are already excellently educated and are currently highly respected as Data Scientists, the new Master program will generate a pool of more interdisciplinary oriented and qualified graduates.
The Data Science Lab has been founded in 2015 in cooperation with the Siemens AG, however, it is open to all prospective partners from industry that share the same visions on Data Science.
49.2 Supporting and Promoting Future Talents in Data Science
A major challenge in facilitating the full potential of Data Science in the corporate as well as in the scientific sector is the availability of young talents which poses the necessary skills to solve practical problems with cutting‐edge data analytics methods.
To understand, apply and develop data‐driven solutions, a profound knowledge in statistics is necessary in order to ensure the correctness of the derived knowledge. Aspects like the quality of available data and the significance of the computed patterns and predictions are mandatory to the usefulness of any data science process.
On the other hand, modern information technology like GPU computing and cloud based systems are key technologies to apply novel prediction techniques and analysis tools. Without the ongoing development in the area of informatics, the recent success stories in machine learning, data analytics and artificial intelligence would not have been possible. These rapid developments allowed for collecting, accessing and integrating data in new dimensions. Furthermore, new technologies to enable new hardware architectures were developed to store distribute and manage data. Finally, new data analytics methods improved the quality of predictions and patterns which are infeasible without massive amounts of data and large scale computational infrastructure.
To provide a training program teaching the key skills from statistics and informatics, the LMU started the new elite program “Master Data Science” in winter 2016. The program is funded by the Bavarian Elite Network and accepts highly skilled international talents having a background in both disciplines. The training program provides the necessary theor
etical background, integrates recent technologies and puts a strong emphasize on the practical application of the acquired skills. Thus, the curriculum includes a practical data science project where the participants solve practical problems of corporate partners. Furthermore, there are summer schools and scientific talks from invited researchers and practitioners on recent developments. The pure analytical skills being acquired during the program are complemented with modules on data ethics, human computation and data security. The program concludes with a master thesis giving the students the opportunity to extend the state of the start when examining their own research project in the area of data science.