Information Technology Ecosystems

Overview

Twenty years ago, data accuracy focused on metadata across one or a few systems, and the analysis was usually local. Information technology ecosystems were few in number and accessible only through a few subject matter experts. In large corporations today, hundreds of systems are stitched together end-to-end with the metadata needed to execute the quote-to-cash and related workflows. In large organizations, there may be several hundred workflows and thousands of process steps, each containing work instructions and other information needed to transform data into usable information. Figure 8.1 describes a generic quote-to-cash ecosystem. Other workstream examples include new product introduction, manufacturing planning, procurement and sourcing, marketing, master data management, sales enablement, reporting of various types, accounting and tax compliance, customer service issues and reverse logistics management, human resources and talent life cycle management, as well as information technology (IT). Additional complexity exists when different tools are used for similar functions (e.g., quoting or ordering in different regions and languages). As an example, if an organization has been acquiring other organizations, legacy IT platforms and applications proliferate. The flow of metadata through the combined IT ecosystem becomes more complicated, and specialized software is required to map and extract it for analysis.

Metadata is information about data fields. Sales and invoicing examples include account name; company name; first name, last name; e-mail

FIGURE 8.1

Quote-to-cash ecosystem.

address; phone number; billing address; shipping address; city, state, zip code, country, province, postal code, Dun & Bradstreet Universal Numbering System (DUNS) Legal ID, DUNS Site ID, payment terms, sales representative; territory name or code, credit limit, and many more. There are literally thousands of metadata fields across the systems in large organizations. Actual data are created, reviewed, updated, and deleted (shortened to CRUD in IT lingo) in metadata fields.

Let us review the quote-to-cash example in Figure 8.1. It is gathered using multiple sources (intakes) within a large IT ecosystem. Salesforce. com (SFDC) is a common intake platform that can be configured to intake quotes and other customer data to create orders that are passed on to other systems. It is a customer relationship management application that provides a single view of a customer across an organization. It also can provide personalized marketing, enable e-commerce, assist customer support, automate routine sales work tasks, run analytics, and more. The data flowing through this application and others complicates our ability to analyze process issues. First, SFDC is a complicated application requiring help from subject matter experts to capture and analyze its stored metadata and data. Second, there may be different instances (i.e., different configurations) of SFDC in an organization. Third, organizations may have other intake applications to capture customer information, which can range from controlled systems to informal ones. In addition, different market segments may have dedicated Internet portals that capture customer or partner information.

The intake tools obtain metadata from other systems such as a customer master data management application. There can only be one master data management system; however, if an organization acquires others or evolves, master data management may exist in several systems and will need to be reconciled. Customer master data include metadata related to DUNS information for sites and legal entity, as well as account name, billing and shipping addresses, contact preferences, and sales-relevant metadata such as sales territory identification, sales representative, and other information. Customer and sales intake information is merged with pricing, credit, service, and entitlement information. Then it is passed to the order management system.

In large organizations, there may be several ordering management systems across different regions. The order will be checked by the order management and revenue teams and scheduled for production in the manufacturing IT systems. Each of these teams works through an IT ecosystem that controls ordering, manufacturing, fulfillment, logistics, returns, and other work. The result is thousands of different metadata fields employed to process the quote-to-cash process. The data complexity increases as additional workstreams are engaged.

How does the IT ecosystem impact process improvement efforts? Let’s use order accuracy as an example. Starting at the highest level, assume we were provided first with a historical baseline of order accuracy (i.e., the percent of orders received by customers without complaint because the order was delivered in the right quantity, quality, and on time). And then we were given a second piece of information, namely that the percentage of order accuracy is decreasing over time. We would begin analyzing (i.e., stratifying) the historical data to identify the major contributors of lower accuracy over time. This type of analysis has become more complicated as the number of systems and subject matter experts needed to complete it increases. Complicating any supply chain analysis is the enormous size of current databases (i.e., Big Data). We will discuss Big Data and its analytics in Chapter 10. Although IT ecosystems have become complicated, process improvement methods are evolving to manage this complexity to gain meaningful insights into the root causes for process issues.

The capability to create and improve processes is enabled, in part, because of the explosive growth of sophisticated diagnostic software designed to acquire data across an IT ecosystem and then analyze it using specialized algorithms. Global collaboration tools also allow the exchange of information between teams on projects that span all global supply chain functions. Technological sophistication has evolved through the major IT functions as shown in Table 8.1. These include the business process management suite (BPMS), business process management (BPM), business process modeling and analysis, business intelligence (BI), business activity monitoring (BAM), enterprise application integration (EAI), and workflow management, as well as Internet transactions, e-mail, and standardized enterprise resource planning (ERP) systems. In this chapter, our goal is to discuss these systems, including their tools, methods, and concepts as they apply to the design and management of global supply chains.

The benefits of successfully implementing IT technology include work task automation and elimination of manual work. In addition to automation, the emphasis has shifted to adding intelligence for monitoring, management, and control of processes. This has been the focus for the

TABLE 8.1

IT Applications

System

Description

• Business Process Management Suite (BPMS)

• An application that integrates several other IT applications to improve system coordination

• Business Process Management (BPM)

• An application that manages a process to coordinate it operations.

• Business Process Modeling and Analysis (BMA)

• Tools that use rules to model workflows and report change sin metrics relative to time, cost and quality to determine an optimum configuration.

• Business Intelligence (BI)

• Using tool to gather data across an IT ecosystem to identify underlying patterns for insight.

• Business Activity Monitoring (BAM)

• Decision rule applied to a model to monitor and control a process or system.

• Enterprise Application Integration (EA1)

* A system that integrates several supply chain functions and their user interfaces (UI) into one UI.

deployment of BAM, BI, business process modeling and analysis, and BPM. Intelligent IT applications promote the sharing and leveraging of information across an organization’s supply chain. IT system modularization also enables greater system configurability and flexibility and promotes global integration and coordination, increasing organizational productivity.

A BPMS integrates and coordinates the diverse applications in an ERP system. Prior to BPMS, applications were scattered across an IT ecosystem. The BPMS enables users to exchange data at several levels within the system to coordinate functions of supply chain management, such as ordering, receipt, fulfillment and shipment of products, inventory transactions, customer transactions, service interactions, accounting and payroll functions, materials planning and purchasing activities, sales and marketing forecasts, and bill of material control. There are other ERP functions that vary by industry. Systems configured in BPMS format can be reconfigured to match modifications to the design of a process as material or information changes. Figure 8.2 shows an example of a BPMS application where operations have been removed and others sequenced differently. A BPMS system links to internal and customer and supplier supporting processes (i.e., interfaces), and these external interfaces remain unchanged when internal processes are modified by business analysts.

Figure 8.3 shows how a BPMS application is integrated into a BPM system with rules, process models, software systems, and manual interfaces. Some BPMS software is highly specific for industries such as financial services or call centers. In these applications, the software will usually be configured based on an organization’s customer-segmentation strategy. Organizational resources may be allocated differentially by these BPMS systems. The rules assigned to each customer segment will usually differ. As an example, in financial service industries, customers may be segmented based on their net worth. One rule may state that highly affluent customers wait no more than ten seconds for their call to be answered versus other customer segments having a small net worth. Also, the type of services provided may vary, perhaps with highly affluent customers being routed to highly trained agents.

A BPMS system can be configured to monitor and control the volume of traffic in a call center. An additional capability may include product and service cross-selling or providing useful information to customers. In some applications, BPMS software must be initially configured by IT, but, once deployed, it can be modified easily by business analysts with minimal IT support. The rules driving BPMS software can be ranged from a few to several hundred, depending on the complexity of the process being monitored and controlled.

BPMS applications use three implementation components. First are manual modifications by its business users. In these situations, work lists can be modified by users to specify different process sequences based on process modifications. A second implementation element could include invoked software applications such as a system forecast. Finally, a third implementation element could be one of several ERP applications. The coordination and execution of BPMS software can be internally managed or web hosted by third-party providers. An important difference between BPMS versus ERP software is that ERP software must be compiled and is somewhat inflexible because its rules cannot be changed without direct IT support.

BPM is an IT management discipline enabling processes to be easily modified, monitored, and controlled. It provides a process-centric approach to workflow management. The business impact of BPM implementation is

FIGURE 8.2

Business process management suite (BPMS). BPMS is a rule engine that manages the business processes.

FIGURE 8.3

Business process management application.

productivity increases between 50% and 80%, a reduction in transition errors between 90% and 95%,. Also, IT investment savings is another benefit because business analysts can reconfigure and analyze processes without IT support. BPM uses loose process coupling. It promotes the reuse of lower-level software. In fact, software functionality is not 100% implemented in an initial BPS application, but rather evolves as software applications are refined by user experience. This reduces software development and maintenance costs from an Agile perspective.

A BPM deployment process begins with determining business modeling specifications and the required user-specified design goals and objectives. The initial focus of a BPM team is to capture the basic required functionality and work down into lower application levels until rules and procedures are set for the process being modeled. Supporting higher-level functionality are middleware applications such as system hardware, peripheral hardware components, software components, and telecommunications equipment and related hardware. These components are integrated through language mapping. Software is used to enable dynamic scheduling and real-time modeling of a process. Supporting this capability are modeling and metadata specifications and data. Other related functionality includes structuring mechanisms for transaction data collection and time management, as well as object location and security management. Additional features include object management group domain specifications, which vary by industry, as well as embedded intelligence and security specifications. A BPM system provides higher-level system capabilities and flexibility through these normalized and standardized processes.

BPM, including simulation, queuing, and liner programming models, have been in use for decades. More recently, however, they gained popularity because of increasing computing power, a greater need for process productivity, and the increasing complexity of the systems being modeled. The first step is to create an electronic version of the process. This is like constructing detailed process maps of material and information flows. Once a process has been mapped, workflow rules are associated to model’s operations with rules for how to transform material and information into mathematical representations of the transformation processes for each operation. These representations are statistical distributions that describe the parameter distributions, including their initial and final states. Once a process is quantified, an analysis is made to estimate costs, lead times, quality levels, capacity, and other system characteristics, such as yield and uptime. The transition state probabilities for the operations, including frequencies and rates, may also be required to construct the model.

Once the model is populated with the necessary information and its rules are established, analysts evaluate alternative processing scenarios to determine the optimal levels for each resource relative to bottlenecks and capacity-constrained resources. Throughput rates and operational costs are also calculated. Simulations can be run to evaluate process changes, from eliminating non-value-add operations to making modifications to the remaining value-add operations. Final modeling activities include the evaluation of alternative solutions.

BI uses data mining methods to search and aggregate information from diverse and disparate databases for analysis. The resultant information could be a simple list of relevant information associated with the questions that prompted the analysis or incorporated into decision-support algorithms to provide information that can be used to answer other questions. BI methods rely on metadata such as part numbers, customer numbers, and key search words and phrases. These correlate one piece of information to another across disparate databases using a common key or metadata field. Increasingly, algorithms mine text in disparate databases without relying on metadata fields or tags for collecting and aggregation data. An example would be typing in phrases that enable Internet searches of specific topics. BI methods improve the efficiency of clustering, classification, and taxonomy-related analytical methods by identifying data patterns that are associated with event outcomes of interest. As an example, if we find a subset of call center agents with higher productivity and customer satisfaction ratings than other agents, we could use BI data- extraction methods to analyze digital recordings of customer calls and extract key phases associated with customer interactions to provide clues as to which behavioral patterns drive higher productivity and quality. BI methods combined with decision support and process-modeling capabilities can greatly enhance the ability to optimally configure and manage processes.

BAM is used for several important business applications, such as monitoring the status of processes in real time for better control and decision making, or understand relationships between process output metrics and the variables that influence them. This enables an organization to drill down to identify the root causes of process variation or poor performance. In these systems, an alarm event usually signals a potential process issue. This is like mistake-proofing efforts in which error conditions are eliminated from a process to prevent defects. If prevention is not possible, then alarms are placed within a process to predict the likelihood of defect based on transformation models that predict output metric or variable performance based on inputs. BAM is enabled through an integrated messaging system. Alternatively, BAM can be used to proactively improve a process.

A prerequisite for BAM implementation is development of a stable process. Once the process is stable, its key metrics are identified and used to identify the status of process operations. Decision rules and support systems are layered on top of the basic process model to show status and abnormal events. The algorithm analyzes process metrics for patterns and trends as well as violations of business rules and constraints. Notifications are sent to predetermined users with recommendations for action. BAM also enables offline analysis of event occurrences using what-if scenario analyses. Examples include changes in capacity and other process modifications to understand costs, lead times, and other performance criteria.

FIGURE 8.4

Enterprise application integration (EAI).

EAI systems integrate several global supply chain functions. Examples include transactions related to purchasing, accounts payables, accounts receivables, production activity control, fulfillment, order management, inventory transactions and material flow, invoicing, financing, and other functions. Figure 8.4 shows that EAI systems coordinate several IT systems using one interface versus previous configurations that employed numerous system-to-system connections. These previous configurations were not able to integrate IT applications and share information across a global supply chain. The high level of integration in modern supply chains is supported by hardware and software standardization and Internet technologies. The flexible configurability of the IT ecosystem is streamlined using EAI integration. Other benefits are intelligent information routing including instructions, middleware that supports messaging to monitor and control system applications, and an ability to manage several processes virtually.

Previous issues that inhibited the implementation of EAI included lower- level incompatibilities between application programming interfaces (API) and conflicting programming models and client APIs among some EAI platforms. Legacy systems to be integrated also made integration difficult because of security clearances or an inability to create new user accounts or support new APIs. To the extent EAI can be deployed within a global supply chain, its benefits include workload balancing, asynchronous messaging, distributed communications, and business process application sharing as well as access to and sharing of disparate databases. EAI systems support BPMS as well as BM A, BI, BMA, and workflow management.

Table 8.2 summarizes the most common global supply chain IT platforms with a description for each one. These systems include enterprise resource planning (ERP), material requirements planning II (MRPII), material requirements planning (MRP), distribution requirements planning, master production scheduling, forecasting systems, capacity requirements planning, manufacturing automation protocol systems, and warehouse management systems, including auxiliary software systems and the advanced shipping notification system that is used to control material flow between suppliers and their customers.

An ERP system is a more sophisticated version of an MRPII system that includes accounting-related information and the resources needed to plan, manufacture, and ship customer orders. ERP systems also have graphical user interfaces. MRPII was an earlier version of ERP in that it had higher functionality than the original MRP. MRPII functionality includes operational and financial data conversion that enables business planning, and it integrates sales and operations planning, master production planning functions, material requirements planning, and capacity planning. The original MRP systems used bill of material, inventory, and master production schedule information to calculate their net requirements for materials and components that were required for manufacturing and supplier orders. These net requirements were offset by material and component lead times. Distribution requirements planning is a system that replenishes inventory at branch locations throughout a distribution network. It uses time-phased order points or similar logic to translate planned orders to suppliers for every item and its location throughout the distribution system. The master requirements planning schedule is a system that uses sale’s forecasts and order book or firm demand, estimated gross capacity, on-hand inventory levels, and other manufacturing planning information to develop a netted manufacturing schedule. The forecasting system uses historical demand and time-series algorithms to predict future demand by forecasting time intervals over a forecasting time horizon.

TABLE 8.2

Global Supply Chain IT Evolution

IT Platform

Description

• Enterprise Resource Planning

A more sophisticated version of the original MRPII system that includes accounting-related information as well as the resources needed to plan, manufacture, and ship customer orders. These systems also are characterized by graphical user interfaces.

• Material Requirements Planning II (MRPII)

A system with higher functionality than material requirements planning. It includes operational and financial data conversion and allows business planning. It integrates functions related to sales and operations planning, master production planning, material requirements planning, and capacity planning.

• Material Requirements Planning

A system that uses bill of material, inventory, and master production schedule information to calculate net requirements for materials and components for manufacturing and suppliers. The requirements are offset by material and component lead times.

• Distribution Requirements Planning

A system that replenishes inventory at branch locations throughout a distribution network using a time-phased order point or other logic for every item and location to translate planned orders via MRPII to suppliers.

• Master Production Schedule

A system that uses the sales forecast and order book demand, gross capacity, and on-hand inventory manufacturing planning to develop a “netted” manufacturing schedule.

• Forecasting System

A system that uses historical demand and time-series algorithms to predict future demand by forecasting time intervals over the forecasting time horizon.

• Capacity Requirements Planning

A system that uses MRPII information related to open and current manufacturing orders as well as routings and time standards to estimate required labor and machine time across facilities.

• Manufacturing Automation Protocol

A system based on the International Standards Organization, which allows communication between systems from different organizations and depends on the International Standards Organizations open systems interconnections standards.

• Warehouse Management System

A system that dynamically manages received materials and components and assigns an inventory storage location. More advanced versions of these systems enable efficient order fulfillment and cycle-counting activities.

• Advanced Shipping Notification

An integrated system that allows customers and suppliers to know all the items making up an order by their pallet and vehicle using barcode scanning. A prerequisite is a supplier certification program and deployment of information technolog)'

A capacity requirement planning system uses MRPII information related to open and current manufacturing orders as well as product routings and time standards to estimate the required labor and equipment across a supply chain and at a local process level. The manufacturing automation protocol system is based on the International Standards Organization’s open systems interconnections standards that enable communication between systems of different organizations.

Warehouse management systems enable distribution centers to dynamically manage materials and component receipts and assign inventory storage locations. Advanced versions of these systems enable efficient order fulfillment and cycle counting. Finally, advanced shipment notification systems provide customers and suppliers with visibility relative to the items making up their orders by pallet and vehicle and using barcodescanning systems. A prerequisite to deployment of an advanced shipment notification system is a supplier certification program and an IT deployment to ship and receive orders.

Workflow management focuses on the design and management of processes. Software is used to create virtual representations of processes. Virtual process maps are easy to modify and, once quantified, enable analyses, simulations, and other models. More recently, advanced adaptive systems can optimally reconfigure a system dynamically. Business process management (BPM) systems are used to integrate processes. Agile project management is used to coordinate and manage these projects. In summary, the design, management, and control of processes have evolved from highly manual models to virtual models. Virtual process models enable easy reconfiguration as well as the addition, deletion, and modification of operational and governing business rules. These concepts are listed in Table 8.3.

 
Source
< Prev   CONTENTS   Source   Next >