Brett Holland
Chief Operating Officer
Akoya Inc.
Peoria, Ill.

Nelson Jones
Technical Manager
Caterpillar Inc.
Peoria, Ill.

Cost-driver information for a specific part, showing variables that are out of alignment based on typical conditions.

Cost-driver information for a specific part, showing variables that are out of alignment based on typical conditions.


One of the greatest hurdles design engineers face is creating products within cost parameters that the marketplace dictates. This has never been truer than it is today. But it's difficult to reduce costs without an accurate method for identifying what costs really are.

Engineers need a tool that provides cost transparency. Feature-based cost analytics (FBCA) is one such tool. The software, which is available in beta mode through Akoya Inc., combines data and data-mining algorithms to generate cost models.

Using FBCA, engineers can learn accurate cost implications of various scenarios, and make better decisions. Given five or six key features of a part (size, weight, material, and so forth) users can analyze features, construct should-cost curves, and determine the key drivers that affect costs. FBCA software gives engineers a kind of cost-analysis X-ray vision.

Traditionally, product-design engineering and direct-materials procurement have been separate functions. Engineers would design the products and specify the parts and materials. Then, purchasing would identify suppliers and negotiate prices.

But some companies purchase thousands of items for the manufacture of complex products. Often, related materials are sorted into packages for bidding and cost analysis. Some companies even employ manual methods to approximate the cost of these packages, using as a basis some function of weight, size, type of material, or fabrication method (for example, casting, machining, drilling).

These methods, however, are labor intensive and time consuming. Finding ways to lower costs is nearly impossible when the relevant data resides in stacks of spreadsheets, and expertise is divided among individual purchasing experts.

Engineers are generally unfamiliar with the commercial side of materials, while buyers lack design and engineering knowledge. Companies might identify considerable opportunities for savings if personnel from the two areas only cooperated early in the product-development process.

To accomplish this, procurement personnel need access to engineering as well as financial data. They need a technology platform through which they can analyze a vast number of materials. Product development must involve interdepartmental teams with joint responsibility for meeting productcost targets.

FBCA requires such cooperation between engineering and purchasing. As a result, purchasing receives should-cost figures prior to sending out for bids. Most costing tools examine conditions such as supplier reliability, redundant supply sources, and the like. What these tools have ignored are product design features.

Assume an engineer has designed a cast-iron bracket. Using FBCA to get a should-cost estimate for the new part, he simply transfers a 3D model along with an estimate of its volume to the FBCA tool and queries a similar part category (for example, cast-iron brackets). The software has thousands of cast-iron brackets that have been analyzed for lowest cost by weight, number of drilled holes, casting complexity, and other features.

The software automatically extracts the feature information from the new part's 3D model and provides a cost estimate for the newly designed bracket based on the program's analytics and relevant volume/feature data. FBCA provides information on four or five similar low-priced brackets to be used in the request for quote process. The engineer gets a should-cost estimate for the new bracket in 2 min.

Engineers usually have a general idea which features drive the cost of their components. FBCA provides a means to test these ideas across thousands of parts and quantify perceived cost drivers.

Moving from general notions of what drives costs to quantifiable relationships as a function of specific features creates proactive cost control.

That same bracket, for example, would have one shape for a given structural load if it were made of aluminum and another shape if it were made from cast iron. FBCA provides should-cost estimates for these alternative designs, "on the fly." The conventional method for getting cost estimates requires so much time that engineers resort to making best guesses. With FBCA, engineers can quantify costs for parts early in the concept stage.

Feature-based costing continuously updates component costs, which are reflected in the should-cost estimate. The program lets purchasing compare quotes from vendors to those estimates. Armed with reliable information relevant to the cost drivers, purchasing will know when quotes are unrealistic.

Studies show that 70 to 80% of product costs are established by decisions made in the first part of the development process. Later in the development cycle, concurrent-engineering tools that evaluate each processing step help refine the design. Cost analysis that previously took months to complete can now be done in weeks or even days.

Companies that embraced concurrent engineering, design for assembly, design for commonality, design for manufacturing, and so forth lacked a "design for cost" tool that could be used early in the concept stage (even prior to a detailed PRO/E model or labor-intensive design for manufacturing applications).

The object of FBCA is to make front-end processes faster and more accurate. A true shouldcost view of products helps manufacturers better negotiate prices with suppliers and bring new products to market at lower cost and in less time.

During the design phase, FBCA helps companies build cost control into their products. Using 3D drawings, a team can review the original design and suggest modifications that lower production costs without sacrificing quality or function. This tool can be deployed in an analysis platform that turns a company's expertise into standardized rules — sort of like having a "brain in a box."

Key elements of a Feature-Based Cost Analytics solution

  • In-depth visibility into parts data, down to the specific feature level
  • High accuracy of calculations and scenarios with critical insights into root causes of cost misalignments and identification of costreduction opportunities
  • Fast execution of cost modeling and analysis of large volumes of parts to streamline front-end processes, shorten productdevelopment time, and speed up time to market
  • Common portal to provide a common language and unified platform through which design, engineering, and purchasing can collaborate in real time
  • Ease of use through intuitive navigation, convenient access to information and analysis functions, and a flexible user interface
  • Fully scalable to accommodate growth and adapt as needs change
  • Evolving value over time through the incorporation of tribal knowledge and experience, accurate historic records, standardization of product features and attributes, and statistical analysis for dynamic supplier-contract management