Feature-Based Cost Analytics

Routinely harass part suppliers with demands for across-the-board price cuts? Not Peoria, Ill.-based Caterpillar Inc.

Instead of beating up on suppliers with unfocused demands, two Caterpillar managers, Nelson Jones and Syamala Srinivasan, pioneered an alternative, a unique part analysis software tool that quickly and easily recognizes cost-reduction opportunities in a design configuration. The goal, as they saw it, was to automate the identification of logical opportunities for cost-reduction collaboration.

The good news is that not only were Jones and Srinivasan successful, but now anybody can implement the cost-analysis solution. The technology was spun off in 2003 and is now commercially available.

The Peoria, Ill.-based startup Akoya Inc. now has Caterpillar as its largest customer, says Brett Holland, chief operating officer and co-founder.

Brett Holland: "With feature-based cost analytics, companies can build cost control into their products.What inspired Holland to take on the commercialization challenge was hearing the encouraging classroom observation of Joe Sandor, then a professor at the University of Chicago, who said: "What's lacking in the world is an automated way to do feature-based analytics against a given company's purchase order files and CAD drawings."

Sandor, now a Michigan State University professor of supply-chain management, also reminded his student of the drawback of manual methods. (Although an Excel spreadsheet can correlate features to costs, the process is too laborious and restrictive.)

Sandor's conclusion: "The world was waiting for the Akoya solution, a software capability that routinely allows buyers to look at large and complicated bills of material and make sense of the parts' cost relationships to one another. It's one of the most compelling needs of the procurement and supply chain world," says Sandor.

Akoya's software is able to take input from the company's existing files and, through regression analysis, identify pricing and cost trends as they relate to part configurations. Simply put, Akoya's solution can quickly compare the characteristics of thousands of parts with the assumption that similar characteristics should have similar costs. At that scale manual methods might take years.

"If OEM's adopting the software continue to browbeat their suppliers, at least they will browbeat with some specifics that they can both agree on and are actionable," jests Sandor.

Consider feature-based cost analytics a tool for both purchasing and engineering, adds Sandor. "Engineers need to know the cost implications of their design decisions."

Sandor also emphasizes that the solution has applicability broader than purchased parts or new part designs. In addition to sharpening make vs. buy decisions, Akoya's solution opens up savings potential with legacy parts. "Even if your purchasing and parts design processes are leading edge, consider the legacy of parts that don't get a lot of attention once they're in production."

He speculates that Caterpillar, for example, may have more than a half million distinct parts in that category. "Normally companies don't review them. In conventional thinking, engineers just assume what was done yesterday was right and proceed to today's priorities." Adds Holland: "With feature-based cost analytics, companies can build cost control into their products."