Big data technologies not only serve to draw business-relevant information from mountains of data, they are also the foundation for automating services and for completely new business models. Those who want this need to experiment, advises Victor Schlegel.
Victor Schlegel: For example, we have developed an application together with the Swiss Federal Roads Office FEDRO with which we use the movement data from our mobile communications network to place a virtual sensor every 500 metres along the roads. With these sensors, we measure the average speed of the vehicles. In this way, FEDRO can control the traffic as well as detect traffic jams and analyse how they occur. In the future, this can be used as a basis for better planning of construction sites, targeted alleviation of critical points or implementation of optimum countermeasures in real time. If one were to set up such a closely knit system with real sensors, sunk into the road surface, it would be impossible to pay for. A single sensor costs more than 20,000 francs per year.
No, in all projects we provide only anonymised and aggregated data that does not allow any conclusions to be drawn about an individual person even with the aid of third-party information. This includes periodically changing the IDs of the individual devices and using only average values from multiple users. In this process, we adhere to the conditions of Swiss data protection law and to international best practices. Admittedly this «privacy by design» is not absolute – that is fundamentally impossible – but the effort that would have to be put in to identify somebody is extremely great.
We are currently at the start of a fundamental transformation from a service to an information society. More and more services that had been performed by people until now are being taken over by computers. They man helpdesks, write news reports, control cars, manage work teams and find the best investments or the closest taxi. The example of the smartphone transport service Uber impressively shows how fast data-operated companies can turn an industry upside down.
For Victor Schlegel, until September 2016 Senior Manager Big Data Solutions at Swisscom, Europe is moving too cautiously when it comes to technology adaptation. And the business informatics graduate does have the figures in mind here. As an improvement expert with a Lean Six Sigma Black Belt, he previously drove forward IT service management projects and efficiency improvement projects at T-Systems and Credit Suisse.
1/7 Crowds on the banks of Lake Geneva during the Lake Parade. The height of the blue lines corresponds to the number of ravers who dance at this point. In this way, you can, for example, work out when there are too many people per square metre. Numbers can be reduced in relevant sections of the road in good time.
2/7 What is known as heat maps show how many people are currently located in which place at a specific time. In Zurich city centre on a Saturday afternoon, shoppers can be found en masse in the lower part of Bahnhofstrasse and on Rennweg. With such maps, for example, retail traders can find the ideal location for a branch.
3/7 Virtual sensors in the road use anonymised mobile communications data to permanently record the average speeds on the Swiss motorways. By analysing this data, it is possible to trace the mechanisms by which traffic jams form. In the future, this will be used to prevent traffic gridlock by means of better control.
4/7 Anonymised mobile communications data illustrates how Swiss residents travel by train on a daily basis. The thickness of the lines stands for the number of passengers that travelled between two train stations.
5/7 However, the mobile communications network does not stop at the train station. It also shows where the passengers go after their train journey. Only some of those who travel between Berne (blue) and Zurich (red) actually come from the capital city. And the city on the Limmat is not the final destination of most travellers.
6/7 Who is on the go, and how? If we analyse the movement patterns, such as number of stopping points per day, stretch of road, journey duration, return to a place, number of passengers, etc., it is also possible to differentiate between the individual modes of transport that circulate within the city based on the mobile communications data.
7/7 French (blue) and German (yellow) people rarely cross over the «Rösti ditch» – the border between German- and French-speaking Switzerland. Italians (green) make it across the Alps more frequently. Swiss (red) like to head for the mountains. The mobile communications network knows which nationalities visit which places. The tourism industry can use this data to adjust the offers to the new trends.
Really, big data is just a buzzword. In principle, it’s all about classic data analysis and artificial intelligence methods – but on a magnified scale. This means that, as with every business intelligence project, it is not the technology that is decisive, but the question. Once you have determined the most crucial business problem, you should spend a few days gathering information about the best practices. However, after that I recommend starting with the first trials as soon as possible. Because the earlier you fail, the better! The right questions and the best data routes to the answers can only be found through the fastest possible iteration cycles, in which you can learn one step at a time.
«The platform will turn big data into a service. This means companies will no longer have to establish and operate complicated infrastructures themselves.»
Viktor Schlegel, until September 2016 Senior Manager Big Data Solutions at Swisscom
Providing anonymised and aggregated mobile communications data and our analytics expertise is just the first step. For us, big data is much more extensive. We are in the process of setting up a platform on which, in a few years, companies will be able to connect and analyse all manner of internal and external data. This will contain all the relevant public sources, such as weather and GIS (geoinformation system) data or social media, as well as private services that choose to participate. The aforementioned data from the mobile communications network is just a part of this. This will be supplemented with the data that the companies upload themselves, such as from their CRM and logistics systems. We will not only provide them with a highly secure, efficient environment, a vast quantity of third-party data and the best evaluation tools for their analyses, but also with algorithms that they can use to automatically anonymise their own data. The platform will turn big data into a service. This means companies will no longer have to establish and operate complicated infrastructures themselves. However, data sovereignty will remain with the customer at all times. This is one of the basic principles of our strategy. In exactly the same way, we do not give preferential treatment to individual companies through exclusive contracts. In addition, we implement only projects that have also been checked for society’s acceptance – a task that is the responsibility of an ethics council.
Effective analytics is a complex topic. Apart from an efficient hardware and software infrastructure, it requires a great deal of special expertise in various disciplines, such as mathematics, artificial intelligence, data modelling, programming and IT security. Very few companies already have the relevant skills, and the specialists are highly sought after. With our experienced experts, we can already support companies effectively today. In the future, the platform will contain a large part of the special expertise as standardised functionalities.
More and more routine work will be dealt with by machines. People will use their time for intellectual activities. The things we can already observe today will increase exponentially: what are known as elastic companies, such as Amazon, Nike or Google, will be successful. With their analytical capabilities, they recognise the market trends earlier and better. They use these insights to try out new business models. Many companies fail here, but those that succeed are fundamentally changing the rules of play. It is significant that almost all of today’s elastic companies come from the USA. The Swiss need to become more willing to experiment if they are to keep up.
More on the topic