Engineering systems encompass a broad class of real-world systems which go beyond traditional boundaries of engineered artifacts to interact with and contribute to key societal functions. Due to their large scale, long lifetime, and proximity to social systems, their design process evokes complex features such as adaptation, self-organization, and emergence that take place over long timescales. Achieving broader objectives such as sustainability or resource efficiency requires the collective efforts of a diverse set of actors which does not align well with existing system engineering approaches that rely on a central decision-making authority. Albeit desirable, cooperation among entities can prove challenging to overcome different objectives, requirements, policies, and willingness to share resources and align decisions with top-down goals. A better understanding of how the fundamental structure of a design problem (i.e. strategy dynamics) facilitates or inhibits collective action is essential to help improve design processes, mechanisms, and incentives to achieve desired collective results.

The objective of this research is to understand how the strategy dynamics exhibited by the structure of engineering design problems influence cooperative behaviors and collective performance. The intent is not to optimize or otherwise prescribe solutions to multi-actor design problems but to understand inherent trade-offs, risks, and dynamics between individual actors generalizable to identifiable problem classes. Building on game theory literature, we identify and describe how design problems can exhibit four strategy dynamical domains associated with low and high levels of structural fear and greed. Harmony dynamics (low fear, low greed) is most representative of “cooperative” design problems where all actors have aligned objectives. Coexistence dynamics (low fear, high greed) represent “brinksmanship” design problems where actors struggle to obtain power or authority, for example, in a leader-follower design process. Bistability dynamics (high fear, low greed) is most representative of “collaborative” design problems where actors trade upside potential and downside risk arising from mutual dependencies. Finally, defection dynamics (high fear, high greed) represent a “egotistic” design problem where actors trade individual versus collective benefits.

To evaluate the effect of strategy dynamics on outcomes of design activities, this study formulates bi-level parameter design tasks that exhibit static strategy dynamics. Designer pairs explore a lower-level joint parameter design space (each designer controls one integer design parameter, represented as axes on a two-dimensional grid) to search and optimize for individual objectives (values) in two strategic contexts. Next, designers select an upper-level strategy that modifies the resulting lower-level values based on the underlying strategy dynamics. For example, a design task with a harmony dynamic provides mutually highest-value outcomes if both designers ultimately select the same strategy. Iterative design processes permit upper-level strategy dynamics to influence lower-level design and vice-versa. Design outcomes are evaluated based on design efficiency (maximize one designer's potential value), individual efficiency (maximize one designer’s value), collective efficiency (maximize the product of designer values), and equality (minimize the difference of designer values).

To compare the effect of strategy dynamics in design problems to prior results in other domains, a human-subject design experiment conducts 240 bi-level parameter tasks, administered with four volunteers in each of ten sessions and twelve tasks per rotating pair (three tasks from each dynamic). Despite all four dynamics achieving similar lower-level design efficiency, harmony and bistability show highest individual and collective efficiency and equality. Results indicate cognizant actors generally follow non-cooperative single-shot game normative strategies for some strategy dynamics (harmony and coexistence) but not strictly for others (bistability and defection). Cumulative link model regression analysis shows a structural greed factor contributing to strategy dynamics has a stronger effect on collective outcomes compared to a fear factor. This result may be attributed to social interaction and trust between designers, similar to analysis of repeated games and evolutionary game theory. Results of this study establish a foundation for future work to study strategic decision-making in engineering design problems and enable new methods and processes to mitigate potential unfavorable effects of their underlying strategy dynamics through social constructs or mechanism design.


Structured to Succeed? Strategy Dynamics in Engineering Systems Design and Their Effect on Collective Performance

Division / Sponsors: 
Design Engineering Division and Computers & Information In Engineering Division


Paper Proceedings: https://asmedigitalcollection.asme.org/IDETC-CIE/IDETC-CIE2020/volume/83907

IDETC2020 - Virtual Conference

Supporting the need among researchers worldwide for additional online and multimedia content, ASME now offers the ASME Conference Video Collection. This platform will collect and provide long-term access to conference presentation videos, encompassing all technology areas within ASME's technical conference and events program. About 30,000-40,000 engineers across all mechanical engineering and related disciplines participate in ASME conferences each year. This includes authors of approximately 6,000 to 8,000 technical papers published by ASME, the majority of whom also submit video presentations for our virtual conferences. The video collection will be available as an add-on to existing product packages or as a separate subscription.