What to make and how: Data Science and Manufacturing

In 1985, when U.S. semiconductor manufacturers were developing and building the fastest and smartest computer chips, their global competitors were targeting the U.S. semiconductor industry.  The Unites States and its global competitors made different strategic decisions. 

While U.S. companies decided to focus on the next generation of chips, global competitors decided to produce the slower and older generation chips—and do it well. This decision led to the continual erosion of their manufacturing capability, and resulted in the loss of money.

They labeled their global competitors copycats without any intelligence and brainpower. The negative cash flow eventually shut down development work on the newest and fastest chips at U.S. companies, while positive cash flow from the slower chips enabled the global competitors to eventually invest in the fastest chips built on the knowledge generated by U.S. companies.

The essence is to choose between what the customers need today or to manufacture the latest and the greatest.

BUSINESS APPLICATION

WHAT IS MANUFACTURING ANALYTICS

Manufacturing Analytics is the aggregation, analysis, and role-based visualization and reporting of data representing the manufacturing process. Manufacturing Analytics enables the shift from reactive to predictive process management by detecting potential problems before they affect the process, lower quality, and increase costs. Manufacturers can then reach the ultimate goal of a stable, well understood process. Using Manufacturing Analytics will deliver statistics-based process understanding and make KPIs (Key Performance Indicators) more sensitive to their underlying components and the processes they are measuring allowing managers to react more quickly as conditions and results change.

 

WHY ITS NEEDED IN MANUFACTURING

Over the past several years, manufacturers have made significant investments in systems that collect, manage, and report data about their processes. All of these systems, from DCS and SCADA/HMI on the plant floor, to manufacturing execution (MES), laboratory (LIMS), supply chain (SCM) and enterprise (ERP) applications either create, collect, store, use, or interact with data that represents or affects some element of the manufacturing process.

Today’s challenge in manufacturing is not getting information about a process. Rather, the puzzle that manufacturers want to solve is how to get more value from all the data that is already being collected. Their goal is to provide process managers with the information they need to better understand their processes, identify improvement opportunities, and make confident decisions that will improve yields and reduce costs.

Quality improvement programs such as lean manufacturing, Six-Sigma, and TQM depend on the accurate collection and analysis of relevant data. When successful, these programs lead to continuous process improvement, confident decision making, quick response to problems, and significant cost reductions. Manufacturing Analytics is essential to this process.

NEED FOR MANUFACTURING ANALYTICS PROFESSIONALS

Within a manufacturing organization the major challenges are how to shorten the product development life cycle; and how to manage the entire cycle from conception to product retirement. The window of opportunity for most products is shrinking. It is more challenging than ever to launch a new product before a competitor does. However, no company dare cut corners on quality in the effort to reduce production costs and time. Instead, companies need to analyze and simulate the entire development process from component sourcing to final product, including process routes and operations. Manufacturers must figure out how to break down the development process into multiple segments that can be performed in parallel. It is in this effort that manufacturers collaborate with partners to build product components at a lower overall development cost. It is for the same reason that business process optimization (BPO) requires professionals with skills which suit these demands.

EXAMPLE FROM THE INDUSTRY

SAS FOR MANUFACTURING

Manufacturers can no longer rely on low prices, high quality and on-time delivery to keep them on top. These attributes, which were advantages a decade ago, are now the minimum requirements to stay in the game – and the rules are constantly changing. Manufacturers face globalization, more competition than ever, and customers whose demands reflect their own knowledge and expectations of a global market. Today, manufacturers must track and move extensive inventories, generate a greater number of products, negotiate with numerous suppliers, and maintain a multitude of quality standards. They also have an ever-increasing need to acquire, satisfy and retain additional customers to remain profitable. Because of these complex pressures, it is imperative that all links in the supply chain be managed successfully.

 

http://www.sas.com/industry/mfg/index.html

 

CAREER OPTIONS AND HOW ATI CAN HELP

Careers in Manufacturing Analytics

The key job responsibilities for manufacturing analytics professionals may include:

Building and leading a team to support the global manufacturing sites and Finance Planning and Analysis teams. 

Building out and leading a global Manufacturing Analytics team to support plants as well as Finance Planning and Analysis team.

Build statistical models using SAS.

Conducting high level data analysis planning, reconciles vendor provided research, liaison to operations and platformmanagement and determining high level tool and system use.

Job Roles in Manufacturing Analytics

Manufacturing Associate

Manufacturing Process Manager

Manufacturing Specialist

Application Manager, Manufacturing & Quality

Director of Analytics, Manufacturing

How ATI can help

Improve process visualization and understanding

Detect significant events and problems

Identify sources of variation

Communicate key quality information

Support confident decision making

Drive continuous process improvement

Support development of best practices

Move from reactive to predictive process management

Lower costs and increase yield