Capturing Human Sequence-Learning Abilities in Configuration Design Tasks through Markov Chains

Abstract

Design often involves searching for a solution by iteratively modifying and adjusting a current design. Through this process, designers improve the quality of the current design as well as learning what patterns of operations are most likely to lead to the quickest future improvements. Prior work in psychology has shown that humans can be adept at learning how to apply short sequences of operations for maximum effect while solving a problem. This work explores the sequencing of operations specifically within the domain of engineering design by examining the results of two human studies in which participants created solutions to configuration design problems. First, a statistical analysis of the data from those studies uses Markov chains to show that meaningful operation sequences exist, and can be accurately described using first-order models. Second, this work uses an agent-based modeling framework in conjunction with Markov chain concepts to simulate the performance of teams with and without the ability to learn sequences. These computational studies confirm the assumption that sequence-learning abilities are helpful during design.Journal of Mechanical Design 2

Designing Like a Human: Simulating Cognitive Style in Teamwork with an Agent-Based Model

“…The CISAT framework [33,53,42] is another agent-based model which uses contextualized problems to study how problem characteristics affect the optimal team process and team characteristics. The CISAT framework reflects eight characteristics of both team activity and individual cognition, namely: organic interaction timing, quality-informed solution sharing, quality bias re-duction, self-bias, operational learning, breadth versus depth solution search, and satisficing [33].…”

Design Strategy Transfer in Cognitively-Inspired Agents

“…Frequent use of sequential behavior while solving a task can also be indicative of expertise [22]. Markov Chain models have previously been utilized to represent the sequencing of design operations by designers for configuration design problems [8]. In a first-order Markov chain model, relations between the current operation and the immediately previous operation are captured.…”

“…This relational dependency can be represented as a square matrix T which is k x k in size for each of the k different design operations that are defined in the problem. The matrix T and its probabilistic significance in design have been explained by McComb, et al [8]. This Markov Chain model was later introduced in CISAT as a generative operation-selecting mechanism for the computational agents.…”

“…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]. …”