image

Connections

Solutions Corner

Timing is
everything

Teradata Manufacturing Analytics Platform accelerates time to decision, improving operational efficiencies.

Imagine a state-of-the-art device that an engineering team has recently developed and is preparing for the global market. The team simulates upwards of thousands of scenarios to show how the new product will perform. Then, it assembles a few prototypes to test in real-world conditions.

When the first units off the assembly line are tested, only limited comparisons can be made against the performance specifications. This is because simulation data in the engineering database silo has not been linked to the company’s product life cycle management system to understand if the actual performance of the product matched the cost estimates. Consequently, manufacturing continues to test and ship the product at risk, which blindly consumes resources, time and money while data is being combined and analyzed.

Teradata Manufacturing Analytics Platform

Working inside a massively parallel data warehouse from Teradata, the Teradata Manufacturing Analytics Platform transforms enormous amounts of operations data into meaningful information in near real time. Using in-database analytics, the platform gives immediate visibility deep into supply chain data, so vital operational metrics are pinpointed as they happen instead of days or weeks later. With this capability, issues are found and fixed before they escalate into significant and costly problems.

The Teradata Manufacturing Analytics Platform provides an Active Enterprise Intelligence environment to improve operational efficiencies and manufacturing performance by reducing the time-to-decision process. It applies an industry-specific logical data model (LDM) that consolidates manufacturing, engineering, logistics and other types of operational data. The LDM describes the relationships among the data, while the physical data model applies in-database analytics across those relationships.

By leveraging the Teradata massively parallel architecture, the Teradata Manufacturing Analytics Platform eliminates the time-consuming work of pushing data into and out of external sources. It integrates manufacturing analytics with business processes, off-loading computational demands on a company’s business intelligence (BI) infrastructure. The platform also provides an engine to power business-specific analytics across many third-party tools.

—P.D.

image

Success delayed

In business, data about product performance, yields, delivery schedules and a host of other parameters can make or break a product launch, influence stockholder confidence and, ultimately, affect the business itself. The fundamental issue is not so much whether the data is available; it’s the time lag between getting information and turning it into actionable knowledge. While engineering can implement a product design change that adheres to performance specifications, and manufacturing yield numbers can be factored into sales plans, the timing of when that information is available is critical.

The need to immediately make precise, contextual decisions makes shortening the time-to-decision process a strategic goal for any manufacturing business. Yet many organizations that work with multiple systems find that before they can analyze information and subsequently make a rational decision, they need to load the data into yet another dedicated business intelligence (BI) system.

Put time on your side

If a business could eliminate data latency, the time-to-decision process of data analysis could be shortened dramatically. Instead of matching data from various silos in a separate decision-support activity, the key is to consolidate the silos into a single source, providing a centralized relational view for analysis.

What’s it worth?

Product yield is a significant issue that affects the business. A newly manufactured device might fulfill all of the engineering performance specifications, but the yield of the units might not meet business requirements. Tying expectations set by marketing and sales forecast data with operational data for product yield tells the business side what kind of profit the company will make based on current costing data. If there is too great of a time lag in determining how much the production of a device will actually cost (based on yield from the assembly line) compared with what was forecast, the company’s profit margins, pricing strategies, promotion plans and original equipment manufacturer deals can be put in jeopardy.

—P.D.

When a single, high-performance in-database analysis platform is used, data movement between systems is no longer necessary and data latency is minimized. A central repository for engineering, manufacturing, forecasting and other relevant data sources enables analytical processes on the data that, in turn, make operational knowledge immediate and actionable. Field results show that companies can make operational decisions five to 15 times faster after implementing such a system. The larger the data volumes involved, the greater the return on time-to-decision efficiency.

This quicker response time means proven benefits can be found in discovering and fixing design flaws while the product is still in the development stage. Workflow is also vastly improved. More importantly, supply chain key performance indicators signal critical states so appropriate action can be taken immediately. In addition, new kinds of data analysis can be performed to generate knowledge about, for instance, product returns and warranty or claims activities, resulting in better forecasting and decision making.

With a centralized in-database analysis process, operational efficiency will improve. Faster time to value will result in reduced engineering and supply chain problems, and improved bottom-line numbers.


Your Comment:
  
Your Rating:

Comments
Fuzzy Logix