Insights Collaboration Space – Deploying smart digital services together

What It Is All About

Everyone is talking about Data Science and Artificial Intelligence (AI). In connection with  digital ecosystems, both offer great potential for novel business models. Most companies already have a lot of data, so it seems only natural to raise this treasure trove of data. Although this issue now gets very high priority, also at the management level, unfortunately it is often only addressed in a very technical way. For example, all data is sometimes blindly stored in huge data lakes. Or an attempt is made to fully and extensively document and catalog all data sources and data in the company. However, such endeavors are rarely successful or are ever completed in a satisfactory manner. Furthermore, the result is an overwhelming amount of data and data sources. Using such a basis, it is extremely challenging to develop a digital service that offers real added value to the end user. It is rather rare that data scientists can, at the first go, identify such spectacular correlations as the American supermarket chain that was able to predict a teenager’s pregnancy before the teenager’s father knew anything about this pregnancy. In our experience, often only already well-known or obvious correlations are found if this “bottom-up” approach is used.

The Challenge

A much more crucial question is how to approach the topic in a targeted and at the same time agile way. The starting point for all considerations should always be the direct customer benefit. It is therefore important, as is common in Design Thinking, to first develop a deeper understanding of the problem in order to literally “get a feeling” for the problems and challenges actually faced by the customer. Based on this, several hundred ideas can quickly emerge in the context of so-called innovation workshops, where various creativity techniques and exercises are used. Starting with the most promising ideas, the greatest challenge is to quickly and systematically reach an assessment regarding whether the ideas with the highest priority are technically feasible or not. To formalize this process, we speak of “data science significant requirements”, i.e., non-functional requirements that must be fulfilled by data-driven digital services in order to assure that they are technically feasible. Specifically, data science significant requirements are used to specify which data is fundamentally required and in what quality, resolution, frequency, and quantity it must be available so that the statistical methods used can yield satisfactory results or, for example, that neural networks can be sufficiently trained. The concrete design of these non-functional requirements strongly depends on the functionality and the quality requirements placed on the digital service, for instance in connection with prediction accuracy. It is a fact, however, that these can only be elaborated and evaluated in a team and in cooperation with a wide variety of experts (domain experts, data scientists, statisticians, software engineers, lawyers, ..). Software engineering has always been teamwork, but in the area of Data Science and AI, this is even more true.

In order to enable the data-driven service to be developed to generate insights on a high professional level that are actually “valuable“ for end users, the entire team must have both a very high understanding of the domain and pronounced empathy for the actual problems and needs of the end users. This means that the different experts in the team must cooperate in a highly interdisciplinary and interactive manner. And this is the case to a far greater extent than in software engineering projects in the past.

The Support

With Insights Collaboration Space (ICSpace), the John Deere European Technology Innovation Center (ETIC) and Fraunhofer IESE have developed the first collaboration app for interdisciplinary teams working together on data-driven digital services. The app runs in the Cloud and thanks to state-of-the-art web technology, it is easy and intuitive to use. This is also especially true for experts with a less pronounced technical background. The app offers a very simple graphical notation that supports the user in constructing the data processing pipeline conceptually to such an extent that the data science significant requirements can be derived very easily. The result is a so-called “data dependency tree”, which shows the dependencies between the methods to be used (e.g., Machine Learning methods or statistical methods) and the required data (“datasets”). 

The starting point and the root of the data dependency tree is always a specific need of the customer, formulated as a so-called “business question”. This supports the team in focusing on the analysis of technical feasibility. Based on that, the data dependency tree is built up to the lowest level of the data sources. This provides the team with a simple visual indicator that enables them to see whether the fundamental data science significant requirements are fulfilled or not. 

ICSpace explained on video

Privacy warning

With the click on the play button an external video from wwww.youtube.com is loaded and started. Your data is possible transferred and stored to third party. Do not start the video if you disagree. Find more about the youtube privacy statement under the following link: http://www.youtube.com/t/privacy_at_youtube

We use the provider YouTube for the integration of videos. Like most websites, YouTube uses cookies to collect information about the visitors of its website. When you start the video, this could trigger data processing operations. We have no control over that. For more information about privacy at YouTube, please see their privacy policy at: http://www.youtube.com/t/privacy_at_youtube

More information about ICSpace and the collaboration possibilities can be found in the following short video:

The Result

We have been developing ICSpace for about two and a half years now. Right from the start, we used the app in our projects in order to continuously generate valuable feedback for its continuing development. We found that both team members with a technical background and those with a non-technical background were able to work with it equally well. Domain experts and marketing staff often formulate the business questions, including the associated quality requirements. The domain experts usually also provide a first rough draft of the data dependency tree, which is initially strongly oriented towards specialist knowledge. Software engineers then refine this draft with concrete data sources, and document first datasets and steps for extracting the datasets from the data sources. Next, data scientists evaluate different methods on the datasets and document the results. The evaluation of different  alternative ways to implement the data processing is directly supported by the app in a simple manner. Team members can also document open issues directly in the tree and can thus more easily control handovers and coordination within the team. In addition, we have been able to integrate ICSpace very well into our agile development process. The technical feasibility of data-driven services is evaluated in so-called runway sprints. Here, the data dependency tree is built up completely over a few sprints (usually 2-3 sprints with a duration of two weeks each) until it is clear whether technical feasibility is given, respectively whether all data science significant requirements can be fulfilled or not.

In some projects for the development of services that were largely specified by management, this approach also enabled us to very quickly make it clear to management that the projects were still too ambitious at that time. These projects were then stopped for the time being for traceable and documented reasons. The most frequent reason was that required data was not available in sufficient quality and quantity. ICSpace thus also provides great support in the development of long-term strategies in conjunction with the provision and generation of certain data sources and datasets. Furthermore, it provides very good support for making a decision on whether to enter into partnerships with specific data providers with regard to the product roadmap. The fact that we had to temporarily halt some projects may seem a pity at first glance, but it enabled us to save valuable development resources, which would have been spent for nothing if we had only discovered the problems related to technical feasibility much later in the development process. This enabled us to first focus on those features that were actually technically feasible at the time, and which will ideally provide us with the crucial data basis for developing the next generation of our data-driven services. 

Overall, the experience we have gained with ICSpace to date has been positive throughout. We want to make the app available to other companies facing these challenges, too. Please use the contact possibilities offered on our website if you want to learn more about the product. 

Your Benefits

  • Offer space for collaboration to experts with different technical and domain backgrounds. Our Cloud app will convince you with its simple and intuitive UX concepts.
  • Support knowledge dissemination in your team and in your organization and thereby empower your team to deploy smart digital services together.
  • Know your key data assets and the most important data sources in your company.
  • Facilitate communication with management and create transparency with regard to the technical feasibility of...

Our Competencies and Solutions

Registration for the ICSpace Newsletter

News

  • We keep you up to date! Subscribe now to our newsletter.
  • Meet us at the data2day in Ludwigshafen from 23 to 24 October 2019! We will give a presentation about successful interdisciplinary teams in Smart Farming.

  • On 8 and 9 November 2019, we will be at LAND.TECHNIK AgEng in Hannover! There we will present our agile engineering approach for the development of Smart Farming services.
  •