Researchers use Network Statistics to Unlock Effective Product Development


Cambridge, MA - Scientists at The New England Complex Systems Institute (NECSI) analyzed a series of major companies' large-scale product design and development projects, defining statistical properties that impact speed and cost-effectiveness. Available in the July issue of Management Science, the study examines data from designers of vehicles, software, pharmaceuticals and hospitals.

"We're aiming to offer insight into improving networks that are at the heart of industry," said Dr. Dan Braha, NECSI researcher. "Development often involves hundreds of designers working on an intricate set of interconnected tasks. Consequently, managing the network between them is vital."

Though statistical physics has been used to characterize many networks, this study is the first to apply the technique to essential corporate networks like product development. Among its findings, the study notes several factors common to biological and social networks also impact how quickly a product develops--for example, the need to balance incoming and outgoing information flows. Typically, development is impeded by coupling parts of a system that do not need to be joined, as well as by "information bottlenecks" (i.e. parts of the network with high volumes of both incoming and outgoing information).

"These networks can work even when some parts fail, but the information bottlenecks are their Achilles heel," said Dr. Yaneer Bar Yam, NECSI President. "As with other systems we've studied there are dangers in depending too much on a single node in a network.

"Information bottlenecks along with certain types of coupling can cause development to become an endless series of revisions," said Braha. "Informed managers will boost system performance by preventing them."

The study suggests a number of solutions to these problems, including the alteration of network connections and flows, preferential resource allocation, decoupling task activities, task modularization, and reuse of prior solutions.

The Statistical Mechanics of Complex Product Development: Empirical and Analytical Results appears in the July issue of Management Science.


The New England Complex Systems Institute (NECSI) is an independent non-profit organization promoting the science of complex systems, the mathematical study of how parts of a system give rise to its collective behaviors. Based in Cambridge, MA, NECSI conducts original research, education and dissemination of the understanding of complex systems and its application to the improvement of society.

 

 

Phone: 617-547-4100 | Fax: 617-661-7711 | Email: office at necsi.edu

238 Main St. Suite 319 Cambridge, MA USA