SAP Analytics Cloud: Analytics Designer Hackathon – Deloitte Digital Factory Boardroom

This blog post is a submission to the SAP Analytics Cloud: Analytics Designer Hackathon.

Introduction

Following the principle “show me, don’t tell me”, Deloitte Digital Factories offer our clients hands-on best practices on digitalization to leverage the full potential of their supply chain organization. This is demonstrated by building on mainly two foundational concepts: The mandatory end-to-end integration of all use cases and the strict principle of utilizing industry- and enterprise-grade hard- and software wherever possible. The resulting high degree of complexity leads inadvertently to occasional failures in a rapidly changing, agile environment. In addition, the ramp-up process for most use cases is very complex due to critical information being spread across a multitude of systems – most of them without shared access authorization Therefore, monitoring all assets and keeping up their availability is a resource-consuming task.

The solution aggregates process and status data from a multitude of different sources. For enabled assets, IoT data is aggregated through SAP Leonardo IoT and provided via OData service. The central ERP, an OP SAP S/4HANA system, provides information on material and product stocks, while process data is provided by the SAP MII system.

The Deloitte Digital  Factory Boardroom offers the tools to keep track of every use case and their embedded assets as well as performance KPIs – on a global scale. Aggregating data from aforementioned sources, it is able to provide insights for a wide spectrum of roles. Especially the relationship and correlation between events (such as IT infrastructure faults) and KPI changes or failures (such as robot does not work anymore) is made more tangible than ever before.

Application Description & Key Insights

The application concept was built with a design thinking approach. One of the challenges was to operate between the conflicting priorities of UI/UX unification, heterogeneous data and the variety of possible user roles. Based on those factors we made the decision to implement a generalist drill down approach. As a first step, we created the wireframes for the views as well as key visuals.

Figure 1: Wireframe mockups for the drilldown view

One of the central targets was to find intuitive and unintrusive ways to represent data. After laying out the functional and design baseline, the actual implementation within SAP Analytics Cloud was started.

Implementation

To start the implementation, we created two models: one for the overview of all locations benefiting the spatial features and another one based on our Digital Factory data source. After that, we created the Analytics Application that is to become our dashboard and within that we started building the first view. Using the created data models as a basis, we built the widgets for the first view. The development iteration for that was concluded by an acceptance test and feedback loop from stakeholders at the Digital Factory. We continued the process with an iteration for the next view where we moved in a top-down approach, from general to more specific views until we completed all views in scope.

Screenshot 1: Initial screen

Screenshot 2: Map showing all available locations

Screenshot 3: Specify location in the selected region

Screenshot 4: Overview of the location including state of the stations

Screenshot 5: Live view of the smart production floor

The next steps in evolving the application can be characterized in two dimensions: enablement of more data sources to integrate more assets, resp. gain more insights into integrated assets. A further use case would be the integration of various notification channels to propagate alarms and start a concrete action from personnel. In the same sense, we would connect to action interfaces and trigger direct action in backend systems.