TY - JOUR
T1 - Visualization of uncertainty in complex projects
AU - Yu, Zhongyuan
AU - Folds, Dennis
AU - Bernstein, Joshua
AU - Dicuffa, Sophia
AU - Huston, Harrison
AU - Mcdonough, Stephanie
AU - Kibalo, Lauren
AU - Mcdermott, Tom
N1 - Publisher Copyright:
© 2025 The Authors.
PY - 2025
Y1 - 2025
N2 - In a project entitled "The Future of Managing Megaprojects,"the research team prototyped two advanced data visualization concepts to enhance organization and situational awareness in highly complex, highly uncertain projects. In megaprojects, leadership must adopt methods to discover and track project uncertainties and adapt project planning and resources as necessary changes and actions emerge. Traditionally this has relied on leadership intuition. Generative artificial intelligence (AI) can continually observe both qualitative and quantitative data from diverse project sources to surface hidden information. While this cannot predict outcomes, it can aid the intuition of the project leadership with situational awareness of emerging issues and organization of project tactics for response. We prototyped two novel display concepts, a metaphoric display and a project uncertainty dashboard, to test the ability of generative AI to aid in management of large complex projects. Metaphors can serve as effective tools in visualizing abstract and interconnected concepts, fostering a shared vision, and facilitating discussions among team members and stakeholders. The project uncertainty dashboard used a novel uncertainty classification rubric to alert leadership to project issues arising from unknown or unexpected events. The alerts were created by both open and custom Large Language Models (LLMs) that were trained with eight aspects of uncertainty in the rubric. These prototypes offer new approaches for humans to visualize the complexity of megaprojects as a complex systems management tool.
AB - In a project entitled "The Future of Managing Megaprojects,"the research team prototyped two advanced data visualization concepts to enhance organization and situational awareness in highly complex, highly uncertain projects. In megaprojects, leadership must adopt methods to discover and track project uncertainties and adapt project planning and resources as necessary changes and actions emerge. Traditionally this has relied on leadership intuition. Generative artificial intelligence (AI) can continually observe both qualitative and quantitative data from diverse project sources to surface hidden information. While this cannot predict outcomes, it can aid the intuition of the project leadership with situational awareness of emerging issues and organization of project tactics for response. We prototyped two novel display concepts, a metaphoric display and a project uncertainty dashboard, to test the ability of generative AI to aid in management of large complex projects. Metaphors can serve as effective tools in visualizing abstract and interconnected concepts, fostering a shared vision, and facilitating discussions among team members and stakeholders. The project uncertainty dashboard used a novel uncertainty classification rubric to alert leadership to project issues arising from unknown or unexpected events. The alerts were created by both open and custom Large Language Models (LLMs) that were trained with eight aspects of uncertainty in the rubric. These prototypes offer new approaches for humans to visualize the complexity of megaprojects as a complex systems management tool.
KW - AIViz
KW - Data visualization
KW - Generative artificial intelligence (AI)
KW - Megaproject management
KW - Metaphoric display
UR - https://www.scopus.com/pages/publications/105021829896
UR - https://www.scopus.com/pages/publications/105021829896#tab=citedBy
U2 - 10.1016/j.procs.2025.08.177
DO - 10.1016/j.procs.2025.08.177
M3 - Conference article
AN - SCOPUS:105021829896
SN - 1877-0509
VL - 268
SP - 18
EP - 25
JO - Procedia Computer Science
JF - Procedia Computer Science
T2 - 2025 Complex Adaptive Systems, CAS 2025
Y2 - 5 March 2025 through 7 March 2025
ER -