Collaboration enables multiple actors with different objectives to work together and achieve a goal beyond individual capabilities. However, strategic uncertainty from partners' actions introduces a potential for losses under failed collaboration relative to pursuing an independent system. The fundamental tradeoff between high‐value but uncertain outcomes from collaborative systems and lower‐value but more certain outcomes for independent systems induces a bistability strategic dynamic. Actors exhibit different risk attitudes that impact decisions under uncertainty which complicate shared understanding of collaborative dynamics. This paper investigates how risk attitudes affect design and strategy decisions in collaborative systems through the lens of game theory. First, an analytical model studies the effect of differential risk attitudes in a two‐actor problem with stag‐hunting strategic dynamics formulated as single‐ and bi‐level games. Next, a simulation model pairs actors with different risk attitudes in a 29‐game tournament based on a prior behavioral experiment. Results show that outcomes collaborative design problems change based on the risk attitudes of both actors. Results also emphasize that considering conservative lower‐level design options facilitates collaboration by providing risk‐averse actors with a safer solution. By accepting that decision‐making actors are not all risk‐neutral, future work seeks to develop new design methods to strengthen the adoption of efficient collaborative solutions.
Robustness analysis can support the design and operation of large‐scale water infrastructure projects confronting deeply uncertain futures. However, diverse actors, contextual specificities, sectoral interests, and risk attitudes make it difficult to identify an appropriate robustness metric to rank decision alternatives under deep uncertainty. Here, we clarify how methodological choices affect robustness evaluation using the multi‐actor, multi‐sector Inchampalli‐Nagarjuna Sagar water transfer megaproject in Southern India. We compare a suite of water transfer strategies discovered using evolutionary multi‐objective direct policy search (EMODPS), a strategy proposed by regional authorities and the status quo of no water transfer. We stress‐test these strategies across scenarios that capture climatic and socioeconomic uncertainties and rank them using robustness metrics representing sectoral perspectives and priorities of different actors with varying risk attitudes. Results show a considerable impact of metric choices on robustness rankings of strategies, with compromise solution discovered via EMODPS as robust. The no‐transfer strategy results in the worst water supply robustness with an average volumetric deficit of 17% of total historical demands but emerges as a robust alternative for 6 out of 12 combinations of actor‐sectors with high risk aversion. Also, changes in the amplitude of the Indian Summer Monsoon is identified as the most important uncertain factor determining the failure of strategies. Our findings highlight that the selection of robust solutions should be guided by an understanding of how assumed risk attitudes shape stakeholders' perceptions of vulnerabilities. These findings are generalizable to large infrastructure projects with diverse stakeholders and multisectoral impacts.
more » « less- PAR ID:
- 10441986
- Publisher / Repository:
- DOI PREFIX: 10.1029
- Date Published:
- Journal Name:
- Earth's Future
- Volume:
- 11
- Issue:
- 8
- ISSN:
- 2328-4277
- Format(s):
- Medium: X
- Sponsoring Org:
- National Science Foundation
More Like this
-
more » « less
Abstract -
more » « less
Abstract Regional cooperation among water utilities can improve the robustness of urban water supply systems to challenging and deeply uncertain futures. Through coordination mechanisms such as water transfers and regional demand management, water utilities can improve the efficiency of resource allocation and delay the need for new infrastructure investments. Though cooperation provides utilities with the potential for reduced cost strategies for improving the reliability of their services, two important challenges are worthy of careful consideration. First, regional utilities often have to navigate robustness conflicts stemming from potential asymmetries in their risk exposure, demand dynamics, and the availability of supply resources. Second, successful implementation of candidate compromise water portfolios requires that cooperating utilities understand their operational tolerances to deviations from the recommended regional actions (“imperfect implementation”). This study contributes a framework for identifying compromises across regional robustness conflicts and quantifying tolerances to implementation uncertainties. The framework is demonstrated on a system of four interdependent but institutionally independent water utilities in the Research Triangle region of North Carolina that are confronting robustness conflicts given asymmetries in their vulnerabilities to growing demands and hydroclimatic uncertainties. Our findings highlight that seemingly balanced compromise management strategies can yield significant and potentially surprising unintended consequences that could degrade regional cooperation. Moreover, asymmetries in regional robustness can be amplified with modest deviations from their agreed upon actions. Results of this analysis are broadly applicable to water supply regionalization as a global challenge and provide insights for discovering robust compromises and safe operating spaces for multi‐actor water supply systems.
-
more » « less
There is increasing concern over deep uncertainty in the risk analysis field as probabilistic models of uncertainty cannot always be confidently determined or agreed upon for many of our most pressing contemporary risk challenges. This is particularly true in the climate change adaptation field, and has prompted the development of a number of frameworks aiming to characterize system vulnerabilities and identify robust alternatives. One such methodology is robust decision making (RDM), which uses simulation models to assess how strategies perform over many plausible conditions and then identifies and characterizes those where the strategy fails in a process termed scenario discovery. While many of the problems to which RDM has been applied are characterized by multiple objectives, research to date has provided little insight into how treatment of multiple criteria impacts the failure scenarios identified. In this research, we compare different methods for incorporating multiple objectives into the scenario discovery process to evaluate how they impact the resulting failure scenarios. We use the Lake Tana basin in Ethiopia as a case study, where climatic and environmental uncertainties could impact multiple planned water infrastructure projects, and find that failure scenarios may vary depending on the method used to aggregate multiple criteria. Common methods used to convert multiple attributes into a single utility score can obscure connections between failure scenarios and system performance, limiting the information provided to support decision making. Applying scenario discovery over each performance metric separately provides more nuanced information regarding the relative sensitivity of the objectives to different uncertain parameters, leading to clearer insights on measures that could be taken to improve system robustness and areas where additional research might prove useful.
-
Measuring Risk Attitudes for Strategic Decision-Making in a Collaborative Engineering Design ProcessThis paper evaluates a questionnaire-based risk attitude assessment method to quantify individual risk attitudes for strategic, multi-actor design decisions. A lottery-equivalence questionnaire elicits a utility curve for risky payoffs which is fit to a Constant Absolute Risk Aversion (CARA) model. Secondary data from a multi-actor design experiment provides observations of strategic decisions in two-actor design games for validation. 124 participants complete the risk attitude questionnaire and a series of 29 experimental tasks. Assuming participants follow the risk dominance equilibrium selection criterion, a risk-neutral utility function accurately predicts 62.2% of decisions. Incorporating risk attitudes elicited from the questionnaire only increases the accuracy to 63.3% while incorporating risk attitudes inferred from observations increases the accuracy to 77.5%. While participants exhibit differential risk attitudes in design tasks, results show the lottery-equivalent questionnaire does not provide risk attitudes consistent with strategic design decisions. Results support findings that risk in the engineering domain is contextual. This paper concludes that risk attitude is an important factor in understanding strategic decisions in interactive engineering design settings and understanding risk attitudes can help create more efficient design processes.more » « less
-
Policy gradient methods have become popular in multi-agent reinforcement learning, but they suffer from high variance due to the presence of environmental stochasticity and exploring agents (i.e., non-stationarity), which is potentially worsened by the difficulty in credit assignment. As a result, there is a need for a method that is not only capable of efficiently solving the above two problems but also robust enough to solve a variety of tasks. To this end, we propose a new multi-agent policy gradient method, called Robust Local Advantage (ROLA) Actor-Critic. ROLA allows each agent to learn an individual action-value function as a local critic as well as ameliorating environment non-stationarity via a novel centralized training approach based on a centralized critic. By using this local critic, each agent calculates a baseline to reduce variance on its policy gradient estimation, which results in an expected advantage action-value over other agents’ choices that implicitly improves credit assignment. We evaluate ROLA across diverse benchmarks and show its robustness and effectiveness over a number of state-of-the-art multi-agent policy gradient algorithms.more » « less
