Mining Process Heuristics From Designer Action Data via Hidden Markov Models

Abstract

Configuration design problems, characterized by the assembly of components into a final desired solution, are common in engineering design. Various theoretical approaches have been offered for solving configuration type problems, but few studies have examined the approach that humans naturally use to solve such problems. This work applies data-mining techniques to quantitatively study the processes that designers use to solve configuration design problems. The guiding goal is to extract beneficial design process heuristics that are generalizable to the entire class of problems. The extraction of these human problem-solving heuristics is automated through the application of hidden Markov models to the data from two behavioral studies. Results show that designers proceed through four procedural states in solving configuration design problems, roughly transitioning from topology design to shape and parameter design. High-performing designers are distinguished by their opportunistic tuning of parameters early in the process, enabling a more effective and nuanced search for solutions.

Design Strategy Transfer in Cognitively-Inspired Agents

“…Hidden Markov Model were used to mine process heuristics from behavioral data for the design study of a house cooling system problem [7], indicating that designers transition across 4 states. Further, that work demonstrates that the top performing designers follow specific state-sequencing patterns that differ from those of low-performing designers.…”

“…This helps in creating a model that can replicate human designer results in design solving. This work builds on previous work by McComb, et al [7], which uses Hidden Markov Models to represent human design behavior learned using the Baum-Welch algorithm [28]. Offline learning is analogous to how a designer can contemplate the decisions made during a previous design experience and then select the ones with favorable outcomes for future use.…”

“…This work explores the role of design strategies in guiding this task of design generation based on previous knowledge. The term design strategy is used to refer to a policy, plan, or process heuristic [7] that a designer has for sequencing the operations used in solving a design problem. These strategies have been shown to have very distinct characteristics and can also be used to differentiate between high-and low-performance designer behavior.…”

“…These strategies have been shown to have very distinct characteristics and can also be used to differentiate between high-and low-performance designer behavior. Design strategies can be emulated with probabilistic models; McComb, et al [7,8], represent the design generation process as a stochastic operation selection task while Ji, et al [9], model design preference probabilistically to represent design progression. Also, it has been shown that designers who perform well in configuration design tasks have strong spatial cognition abilities [10].…”

Data on the Configuration Design of Internet-Connected Home Cooling Systems by Engineering Students

“…In total, data was collected for 68 participants, and each participant was allowed to perform 50 design actions in solving the configuration design problem. Major results based on the data presented here have been reported separately, including initial behavioral analysis (McComb et al) [1,2] and design pattern assessments via Markovian modeling [3,4]. …”