Proactive Supply Chain Performance Management with Predictive Analytics

Supply Chain Intelligence Web Portal for Predictive Analytics

A predictive analysis solution is most effective when it is pervasive throughout the organization and helps to drive day-to-day decisions across the business with its scale and enterprise-level performance. Furthermore, providing a way to implement comprehensive predictive analysis intuitively enables self-service data mining for users, which in turn enables the business to promptly gain actionable insights.

In order to overcome the shortcomings of the existing BI and PM client tools, the specialized web portal that enables supply chain users to monitor business processes, collaborate, and take actions is designed. It has been successfully implemented as a pilot project in an automotive company. Automotive supply chain is typically very complex, with many organizations, intertwined processes, and multitude of users with different requirements.

The portal represents a single point of access to all relevant information in a personalized and secured manner. Its composite and service-oriented architecture enables inclusion of different PM components and tools (KPIs, dashboards, scorecards, strategy maps, reports, etc.). PM elements can be personalized and adjusted, and information can be filtered just by using a web browser. PM elements can be defined within the portal and also embedded from the external source (OLAP, another application, or spreadsheets) via web services. This information is presented through special analytical web parts. The portal itself can be a provider (via web services or RSS-really simple syndication) to other applications.

All these capabilities make the proposed PM model and the software system extremely flexible and applicable in various supply chain scenarios and different industries. Because they are based on the SCM metamodel and use a standardized metrics they can be used in various business domains. On the other hand, software system architecture enables relatively easy and fast customization and extensibility. For example, existing KPI components can be easily reused many times, and only needed items can be added to web sites and pages. Also, new (custom) lower-level metrics can be defined in the SCM metamodel and realized in the data warehouse, thus allowing company-specific and industry-specific performance measurement.

With PM portal capabilities, supply chain partners and teams can do the following:

(i) use a predefined PM portal template with out-of-the-box modules optimized for access and management of reports, data connections, spreadsheets, and dashboards. Dashboard pages can contain several web parts, each of them showing information from different data sources.

(ii) communicate strategy and monitor its execution at different levels of the supply chain. KPIs status and trend can be tracked using the special KPI and scorecard web parts. KPIs can display information from different data sources (e.g., OLAP cubes or spreadsheets). The portal also supports the concept of strategy maps by providing a specialized module providing a hierarchical view of the KPI measures across levels of the organization by presenting relationships, priorities, and perspectives. Strategy map can be generated automatically, based on a particular scorecard. Each element on the map is highlighted with appropriate color. This enables visual performance tracking in relation to predefined strategy.

(iii) customize and personalize sites, pages, or modules by adding or removing certain web parts and by applying filter web parts. Filters allow dashboards to be personalized by communicating shared parameters amongst web parts on a dashboard. For example, the current user filter web part automatically filters information based on who is logged on to the computer. This is useful for display of personal information such as customer accounts or tasks that is currently assigned to that user.

Figure 7 shows a specialized SCM scorecard for global supply chain performance management.

Figure 7 Supply chain scorecard with predictive KPI.

It is constructed based on top of the OLAP KPIs, which are again based on the SCM process model and metrics. KPI are created by SCM segments (plan, source, make, deliver, and return) as hierarchies, so it is possible to perform drill-down analysis, track performance against defined goals, and get future performance values and trends.

The presented solution is very flexible in terms of presenting the KPIs. Owing to several specific BI web parts, the portal can display KPIs from the OLAP server, spreadsheets, and other sources (portals, report servers, etc.).

The dashboard page can display numerous metrics and views business on a single screen. The portal supports quick deployment of dashboards assembled from web parts. Each web part can contain a particular view or metric, and users can customize their individual dashboards to display the views that are most meaningful to them, such as those with the metrics they need to monitor on a daily basis.

Additionally, portal supports events and automatic alerting. Users can subscribe to specific documents or keywords and categories, to be notified (via email, SMS, or web feed) when metrics are updated or new intelligence becomes available. They can also use other features, such as planning, enterprise search, subscription, and routing functions, to work with team members on a single item (i.e., scorecard, KPI, etc.) and to automate collaborative performance management processes. The portal also provides fine-grained authentication and authorization which enable secure access and content personalization.

PM web portal enables business users to define and use scorecards and key performance indicators to drive accountability and alignment across the supply chain. Scorecards and KPIs reflect planning, budgeting, and forecasting changes in real time to help users understand the business drivers, challenges, and opportunities they face. Monitoring becomes a part of the regular, day-to-day management process.

The proposed PM model and the software system are successfully used in the automotive company for modeling and design of the supply chain PM system. The system was used by many different users (low-level, midlevel, and top-level managers) for monitoring performance and making KPI predictions and collaborative decision making. The users reported a high level of satisfaction highlighting the following features and benefits: global performance view, standardized metrics, proactive management through predictions, high data quality, on-demand information and services, easy to use and intuitive web-based interface, personalization, customization, and collaboration services.