About This File
The Continuous Supply Chain Accelerator allows users to evaluate historical sales to generate inventory ordering models based on Economic Order Quantity and Safety Stock principles. It also provides tools to optimize allocation of stores to distribution centers based on constraints, as well as generating real-time routing for deliveries using integration with TIBCO Geoanalytics.
Here's a video showing the real-time delivery tracking in action.
Business Scenario
Today, everything is connected and every participant in a global supply chain must access data. So it is essential to lower the barrier between artificial intelligence and human intelligence. With open source at the core and democratizing business intelligence through self-service, an intelligent nervous system is now available to anyone. This augmented intelligence enables a shift from reactive to proactive management of all supply chain areas. Digital twins allow us to predict future system states, anticipate problems, model alternative scenarios and choose an optimal solution. Humans better understand that digital fabric and are able to act in real time.
Typically, planning is the most data-driven process in the supply chain, using a wide range of inputs from Enterprise Resource Planning (ERP) and Supply Chain Management (SCM) planning tools. There is now significant potential to truly redefine the planning process to sense and respond to billions of events a day, in collaboration with suppliers, to make real-time demand and supply adaptation a reality.
Transportation firms have used analytics to improve operations for years to optimize routing and reduce wait times. But most existing analytics are based on historical data, and new possibilities help companies monitor and respond to changing conditions in real-time data from connected land, air and sea vehicles, shipping environment sensors, real-time order flow, supply chain geoanalytics, live traffic patterns and continuous weather forecasting and the rescoring of predictive models.
Concepts
Benefits and Business Value
Making a supply chain more real-time gives business the ability to be more agile and react to competitive and market pressures. Being able to make changes quickly and innovate through a phased implementation approach can deliver near-term value by leveraging existing ERP and SCM infrastructure and tools, and forms the foundation for future projects.
Supply chains are constantly in motion so the first phase of supply chain nervous system adoption is obtaining a 360-degree streaming or near real-time view of the data that impacts supply chain assumptions and forecasts. Real-time analytics and simulation tools provide streaming or near real-time insight to stakeholders to any element that can impact the supply chain, including orders, package scans, inventory updates, in real time. Predictive data science models can be scored with streaming data science against this real-time feed of data and explored by supply chain management experts.
Virtualized data and real-time visibility is just the start. The next phase introduces key elements for scale: dynamic learning, data curation and automation. Dynamic learning is the secret sauce of supply chain innovation. Algorithms applied to streaming data yield smarter supply chain decisions and situational awareness. This algorithmic awareness is the pinnacle of supply chain innovation power. Data creation introduces a culture of curation to metadata management to trace lineage and manage assets and analytics assumptions. With real time analytics in place, automation with streaming data can begin. The best place to start is to automate insights that business users can use to better empower them to see and act on changing factors that impact the supply chain.
The scaling phase of this nervous system focuses on how to scale the center of excellence, enterprise architecture and cloud-hybrid architecture, and edge computing. Enterprise scaling is outside the scope of this paper, but considered insofar as the technology innovations below are expressly designed to future-proof the evolution of data, automation and AI in a global enterprise.
Technical Scenario
The accelerator includes a demonstration called Distribution Logistics. It consists of a series of static Spotfire DXP analyses that evaluate historical sales to build models for ordering. It provides impact analysis for promotions and forecasting for next month sales. Once this forecasted demand is known for each store, an optimization model allocates retail stores to distribution centers using constraints such as maximum capacity and minimizing total distance driven.
The real-time component of the accelerator tracks actual unit sales in stores and triggeres automated re-ordering once safety stock thresholds are reached. Orders are sent to the allocated distribution center. At the start of the day the system will take all orders for each distribution center and build a series of delivery routes based on constraints such as vehicle size and minimizing distance travelled. It then tracks the vehicle deliveries to ensure on-time performance.