DSS for SCM

Supply chain management problems are not rigid and require not just computer but human knowledge to effectively manage the systems. DSS systems are meant to help ensure that analysis is easily understood with the help of computers. DSS in supply chain management is often called Advanced Planning and Scheduling (APS) systems. The systems help create solutions in the following areas:

i. Demand planning- determine accurate forecasts based on historical data, help understand buying patterns of customers, helps facilitate collaboration between suppliers and customers

ii. Supply planning- these are sometimes known as Distribution Resource Planning (DRP) and help in inventory planning, transportation planning, procurement planning, strategic supply chain planning

iii. Manufacturing planning and scheduling- these incorporate the traditional Material Requirement Planning (MRP) system. It helps to efficiently allocate manufacturing resources to meet demand. It can also quote lead times to customers

Three major components of DSS:

i. Input data-this is a database which has the basic information required for decision making. It can come in the form of a data warehouse where all the company’s past transaction are stored; distributed databases that are accessed through a network or a PC-based database extract used for a specific problem.

ii. Analytical tools- Cost calculators, artificial intelligence, simulation, flow analysis, operations research are a few of the tools used.

iii. Presentation tools- these display the results of the data analysis. Data visualization techniques are used to help the user understand the output data

A - Input Data

Information technology such as point-of-sale, ERP, bar coding and electronic commerce provides companies with large amounts of data. The supply chain network requires both static and random data from all various parts of the organization. Static data will include such information as plant locations, warehouses, plant production rates and transportation costs. Dynamic data will encompass such things as forecasts, orders, and current deliveries. The quality of all data needs to be evaluated to ensure that the data required is appropriate and not redundant. Below is an illustration of the type of data a company would need to collect for their logistics network design:

Component	Data
Manufacturer	Location, production capacity, production costs, transportation costs to warehouse
Warehouse	Location, fixed costs, variable costs, inventory turnover, transportation costs to retailers
Retailer	Location, demand for product annually
Product	Volume, weight, holding costs

B- Analytical Tools

Once data has been collected it needs to be analyzed and tools that may be used to do this are as follows:

Queries- Decision makers ask questions about the data

Statistical analysis- this is used to determine trends and patterns

Data mining- these look for hidden patterns, trends, and relationships in data.

Online analytical processing (OLAP) tools- it allows the user to navigate through the hierarchies and dimensions by drilling down. Statistical tools are used to analyze data. This also has presentation tools that will present data after it has been analyzed.

Calculators- these calculate specialized calculations, e.g. accounting costs.

Simulation- this creates a model of the process. The random elements are specified using a probability distribution and each time a random event occurs. The computer will use this to determine what would happen in that specific situation. As the model is running, statistical data is collected and analyzed and then statistical techniques help determine the average outcome and variability of this outcome

Artificial Intelligence- these can be databases of rules that are collected from experts which can be used for specific problems or online intelligent agents. These intelligent agents can be characterized by the number of activities allocated to them, the level of interaction with other intelligent agents and the level of knowledge embedded in it.

Mathematical models and algorithms- Algorithms come in two forms, which are exact algorithms and heuristics. Exact algorithms find the best mathematical solution, and are long to run especially when the problem being solved is complex. Heuristics give good but not optimal solutions. They provide quickly a good solution, unlike exact algorithms.

Factors to consider when determining which analytical tools to use:

. Type of problem being solved

. The required accuracy of the solution

. Problem complexity

. The number and type of quantifiable output measures

. Required DSS speed- for lead times the a faster speed is required

. The number of goals and objectives allocated expected of the decision-maker

The table below shows analytical tools and the problems they are best at solving:

Problem	Tools
Marketing	Query, statistics, data mining
Routing	Heuristics, exact algorithms
Production scheduling	Simulation, heuristics dispatch rules
Logistics network configuration	Simulation, heuristics, exact algorithms
Mode selection	Heuristics, exact algorithms

B- Presentation tools

There are various formats used to present data to the user such as: reports, charts, spreadsheet tables, animation, specialized graphic formats and Geographic Information Systems (GIS).

GIS is the main presentation format used in supply chain DSS. It is an integrated spatial database management and integrated mapping system that allows for management, analysis, retrieval, storage, and display of geographically referenced data. It has the following capabilities:

i. mapping and thematic mapping

ii. buffering/polygon overlay

iii. geo-coding

iv. spatial data analysis

v. geographic data manipulation

vi. interactive data query

vii. database management

viii. mapping and thematic mapping

The fact that GIS can combine the entire top attributes, makes it the best choice for use in DSS in supply chain management. GIS is used to in the following in SCM:

. Routing

. Site selection

. Network analysis

. SCM

GIS may not be as useful in other countries outside the USA, as data may not be available in these countries.

DSS in the USA is usually based on Topologically Integrated Geographic Encoding and Referencing (TIGER)/line files.

Integrating Algorithms and GIS:

When GIS is integrated with mathematical models and algorithms, they give a schematic presentation. GIS supplies the geographic data, whilst standard databases provide the attribute data such as costs, production, and demand information.

Examples:

Example A: Amoco Chemical Corporation

The organization introduced a DSS system that models its multi-echelon logistics, network, objectives and costs. Optimization and simulation are used for analysis. Simulation helps to test the stocking policy, customer service, and other associated costs, once inventory targets have been determined by optimization. There is now a better understanding of inventory costs and shortage costs. Planning, coordination and communication have also improved in the company.

Example B: Nestle

Nestle has signed a contract with SAP to purchase $200 million worth of software that will be accessed by all its employees worldwide. The applications will include internal and external in the following areas:

. E-commerce

. Product life cycles

. Financial and cost management

. Marketing

. Customer relationships

. Knowledge management

Each employee however will only have access to the information required for his/her job. The company hopes that this new software will help them focus on customization of user needs without sacrificing group-wide standardization.

Nestle USA has also signed up IBM Corp to build its direct-to-customer B2B website, nestleezorder.com. The website will allow its customers to place orders and inquire about the status of their orders. Customers will have access to 100 brands and 700 products. The company hopes that this website will support 2,500 customers.

Example C: Autozone

Autozone Inc. a chain of auto parts and accessories stores has invested in vendor-managed inventory programs. The company is now sharing with its suppliers supply chain data, point-of sale data and OLAP applications through the web. Autozone has now managed to lower inventory, achieve faster inventory turnover, gain more customers and make processes more efficient.